diff --git a/docs/notebooks/survival_model.ipynb b/docs/notebooks/survival_model.ipynb index 7ec09a2ad..ba65d54a3 100644 --- a/docs/notebooks/survival_model.ipynb +++ b/docs/notebooks/survival_model.ipynb @@ -17,9 +17,9 @@ "\n", "Sometimes the right way to model discrete, countable events is to model not the counts themselves but rather the **time between events**. This gives us information regarding the rate of an event. Survival models are models for countable things, but the outcomes we want to predict are durations. Durations are continuous deviations from some point of reference (so they are all positive values). \n", "\n", - "The tricky part with survival models is not the probability distribution assigned to the durations, but dealing with censoring. Censoring occurs when the event of interest does not occur in the window of observation. In a simple scenario, this can happen because the observation period ends before the event occurred. Censored individuas (or units) can not just be dropped from the sample. Imagine a cohort of 100 cats who start waiting for adoption at the same time. After one month, half of them have been adopted. Now what is the rate of adoption? You can’t compute it using only the cats who have been adopted. You need to also account for the cats who haven’t yet been adopted. The cats who haven’t been adopted yet, but eventually will be adopted, clearly have longer waiting times than the cats who have already been adopted. So the average rate among those who are already adopted is biased upwards—it is confounded by conditioning on adoption.\n", + "The tricky part with survival models is not the probability distribution assigned to the durations, but dealing with censoring. Censoring occurs when the event of interest does not occur in the window of observation. In a simple scenario, this can happen because the observation period ends before the event occurred. Censored individuals (or units) can not just be dropped from the sample. As an example, we use Richard McElreath's cat adoption example from chapter 11.4 of Statistical Rethinking: Imagine a cohort of 100 cats who start waiting for adoption at the same time. After one month, half of them have been adopted. Now what is the rate of adoption? You can’t compute it using only the cats who have been adopted. You need to also account for the cats who haven’t yet been adopted. The cats who haven’t been adopted yet, but eventually will be adopted, clearly have longer waiting times than the cats who have already been adopted. So the average rate among those who are already adopted is biased upwards—it is confounded by conditioning on adoption.\n", "\n", - "Including censored observations requires a new type of model. The key idea is that the same distribution assumption for the outcome tells us both the probability of any observed duration that end in the event as well as the probability that we would wait the observed duration without seeing the event. For each unit, we assume there is a true _survival time_ $T$ as well as a true censoring time $C$. The survival time represents the time at which the event of interest occurs. The censoring time is the time at which censoring occurs. We observe either: the survival, or the censoring time:\n", + "Including censored observations requires a new type of model. The key idea is that the same distribution assumption for the outcome tells us both the probability of any observed duration that end in the event as well as the probability that we would wait the observed duration without seeing the event. For each unit, we assume there is a true _survival time_ $T$ as well as a true _censoring time_ $C$. The survival time represents the time at which the event of interest occurs. The censoring time is the time at which censoring occurs. We observe either: the survival, or the censoring time:\n", "\n", "$$Y = \\text{min}(T, C)$$\n", "\n", @@ -32,7 +32,7 @@ "source": [ "## Left and right censoring\n", "\n", - "There are two main \"types\" of censoring: right and left. Right sensoring occurs when $T \\ge Y$, i.e. the true event time $T$ is at least as large as the observed time $Y$. This is a consequence of $Y = \\text{min}(T, C)$. Right censoring derives its name from the notion that time is typically read and displayed from left to right. Left sensoring occurs when the true event time $T$ is less than or equal to the observed time $Y$. An example of left censoring could be in a study of pregnancy duration. Suppose that patients are surveyed 250 days (8.2 months) after conception. Some patients may have already had their babies. For these patients, pregnancy duration is **less than** 250 days." + "There are two main \"types\" of censoring: right and left. Right censoring occurs when $T \\ge Y$, i.e. the true event time $T$ is at least as large as the observed time $Y$. This is a consequence of $Y = \\text{min}(T, C)$. Right censoring derives its name from the notion that time is typically read and displayed from left to right. Left sensoring occurs when the true event time $T$ is less than or equal to the observed time $Y$. An example of left censoring could be in a study of pregnancy duration. Suppose that patients are surveyed 250 days (8.2 months) after conception. Some patients may have already had their babies. For these patients, pregnancy duration is **less than** 250 days." ] }, { @@ -65,53 +65,51 @@ "\n", "$$\\hat{S}(d_k) = \\prod_{j=1}^k \\frac{r_j - q_j}{r_j}$$\n", "\n", - "where $\\hat{S}(d_k)$ represents the estimated survival probability up to time $d_k$. The product is taken over all time points up to $k$, where an event occurred. The variables $r_j$ and $q_j$ denote the number of subjects at risk and the number of events at time $d_j$, respectively. The term $\\frac{r_j - q_j}{r_j}$ is the conditional probability of surviving the $j$-th time point given that an individual has survived just before $d_j$. Specifically, $r_j - q_j$ are the number of individuals who survived just before $d_j$ and $r_j$ is the number of individuals who survived just after $d_j$, and $r_j$ are those who were at risk $d_j$.\n", + "where $\\hat{S}(d_k)$ represents the estimated survival probability up to time $d_k$. The product is taken over all time points up to $k$, where an event occurred. The variables $r_j$ and $q_j$ denote the number of subjects at risk and the number of events at time $d_j$, respectively. The term $\\frac{r_j - q_j}{r_j}$ is the conditional probability of surviving the $j$-th time point given that an individual has survived just before $d_j$. Specifically, $r_j - q_j$ are the number of individuals who survived just before $d_j$ and $r_j$ is the number of individuals who survived just after $d_j$, and $r_j$ are those who were at risk $d_j$." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Cat adoption survival function\n", "\n", - "Below is a Python function implementing the Kaplan-Meier estimator to estimate the survival function for cat adoptions." + "Below we use the `KaplanMeierFitter` class of the [lifelines](https://github.com/CamDavidsonPilon/lifelines) package to compute and visualize the survival curve for cat adoptions from an animal shelter in Austin, Texas beginning October 1st, 2013 until May 30th, 2018 (the last day the shelter rescued a cat). The dataset comes from the [City of Austin Open Data Portal](https://data.austintexas.gov/widgets/9t4d-g238) and contains columns such as animal name, date of birth, species, and many more. However, for the purpose of this notebook we are interested in the following columns:\n", + "- `days_to_event` - number of days until the cat was adopted (`date_in` - `date_out`)\n", + "- `out_event` - the reason for the cat leaving this particular shelter, e.g. adopted or transfered.\n", + "- `color` - the color of the cat, e.g. white, blue, brown tabby, black." ] }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 31, "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "WARNING (pytensor.tensor.blas): Using NumPy C-API based implementation for BLAS functions.\n" - ] - } - ], + "outputs": [], "source": [ "import arviz as az\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pandas as pd\n", - "import pymc as pm\n", "import scipy\n", "\n", + "from lifelines import KaplanMeierFitter\n", + "\n", "import bambi as bmb" ] }, { "cell_type": "code", - "execution_count": 2, + "execution_count": 63, "metadata": {}, "outputs": [], "source": [ "url = \"https://raw.githubusercontent.com/rmcelreath/rethinking/master/data/AustinCats.csv\"\n", - "cats = pd.read_csv(url, sep=\";\")\n", - "\n", - "cats_new = cats.copy()\n", - "cats_new[\"adopt\"] = np.where(cats_new[\"out_event\"] == \"Adoption\", \"right\", \"none\")\n", - "cats_new[\"color_id\"] = np.where(cats_new[\"color\"] == \"Black\", 1, 0)\n", - "cats_new = cats_new[[\"days_to_event\", \"adopt\", \"color_id\"]]" + "cats_df = pd.read_csv(url, sep=\";\")" ] }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 64, "metadata": {}, "outputs": [ { @@ -127,7 +125,7 @@ ], "source": [ "plt.figure(figsize=(7, 3))\n", - "plt.hist(cats_new[\"days_to_event\"], bins=250, label=\"Uncensored data\")\n", + "plt.hist(cats_df[\"days_to_event\"], bins=250, label=\"Uncensored data\")\n", "plt.xlim(0, 186) # limit to 6 months for visibility\n", "plt.title(\"Days Until Adoption\")\n", "plt.ylabel(\"Count\")\n", @@ -139,71 +137,30 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The distribution of days until adoption exhibits a long tail with most cats (if we observe the adopt event) being adopted within the first month of inception. Note that the plot has been truncated to six months for better visibility." + "The distribution of days until adoption exhibits a long tail with most cats (if we observe the adopt event) being adopted within the first month of inception. Note that the plot has been truncated to six months for better visibility. Below, we estimate the survival function using the `KaplanMeierFitter` class from the `lifelines` package." ] }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 66, "metadata": {}, "outputs": [], "source": [ - "def kaplan_meier_estimator(d_k, d, times):\n", - " \"\"\"\n", - " Computes the Kaplan-Meier estimator.\n", - "\n", - " Parameters:\n", - " -----------\n", - " - d_k: np.ndarray\n", - " Array of unique observed event times.\n", - " - d: np.ndarray\n", - " Array indicating if the event occurred at the time from the 'd_k' list;\n", - " True if event occurred, False if censored.\n", - " - times: np.ndarray\n", - " Array of times at which the survival function should be evaluated.\n", - "\n", - " Returns:\n", - " --------\n", - " - List of Kaplan-Meier estimates at the specified 'times'.\n", - " \"\"\"\n", - "\n", - " n = len(d_k)\n", - " survival = []\n", - " s_prev = 1\n", - "\n", - " for time in times:\n", - " product = s_prev\n", - " for i in range(n):\n", - " if d_k[i] <= time:\n", - " product *= (1 - d[i] / (n - i))\n", - " survival.append(product)\n", - " s_prev = product\n", - "\n", - " return survival" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "observed_adoption_times = cats_new[cats_new[\"adopt\"] == \"right\"][\"days_to_event\"].values\n", - "adoption = (cats_new[\"adopt\"] == \"right\").values\n", - "evaluation_times = np.arange(0, max(cats_new[\"days_to_event\"]), 1)\n", - "adoption_km_estimates = kaplan_meier_estimator(\n", - " observed_adoption_times, adoption, evaluation_times\n", + "km = KaplanMeierFitter()\n", + "km_adoptions = km.fit(\n", + " cats_df[\"days_to_event\"], \n", + " cats_df[\"out_event\"].apply(lambda x: 1 if x == \"Adoption\" else 0)\n", ")" ] }, { "cell_type": "code", - "execution_count": 40, + "execution_count": 67, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -213,21 +170,20 @@ } ], "source": [ - "plt.figure(figsize=(7, 3))\n", - "plt.step(evaluation_times, adoption_km_estimates, where=\"post\", label=\"Kaplan-Meier Estimate\")\n", - "plt.xlabel(\"Days\")\n", - "plt.ylabel(\"Cat Adoption Probability\")\n", - "plt.title(\"Probability of a Cat Not Being Adopted by time $d_k$\")\n", - "plt.xlim(0, 30)\n", - "plt.legend()\n", - "plt.show()" + "fig, ax = plt.subplots(figsize=(7, 3))\n", + "km_adoptions.plot(label=\"Kaplan-Meier Estimator\", ax=ax)\n", + "ax.set_ylabel(\"Probability of Adoption\")\n", + "ax.set_xlabel(\"Days\")\n", + "ax.set_xlim(0, 365)\n", + "ax.grid(True)\n", + "ax.set_title(\"Cat Adoption Survival Curve\");" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "The Kaplan-Meier estimator shows that by the fifth day, the probability of a cat not being adopted is about $0.15$ percent. By the 30th day, the probability of a cat not being adopted has fallen to nearly $0$. In the sections below, we will discuss how to perform survival analysis using parametric models." + "The Kaplan-Meier estimator shows that by 100 days, the probability of a cat not being adopted is about $0.15$ percent. After 100 days, the probability of cat not being adopted decreases, albeit at a much slower rate. Thus, if a cat hasn't been adopted by the 100th day, it is more likely the cat will continue to wait for adoption. In the next section, we discuss `pm.Censored`, a PyMC distrbution that allows us to model censored data." ] }, { @@ -239,16 +195,18 @@ "The [censored distribution](https://www.pymc.io/projects/docs/en/latest/api/distributions/censored.html) from PyMC allows us to make use of a sequential construction, similar to the Kaplan-Meier estimator outlined above, to model censored data. To understand the `pm.Censored` distribution, lets reason how a distribution may be used to model censored data. For observed adoptions, the probability of observed waiting time can be distributed according to an exponential with some rate $\\lambda$\n", "$$D_i \\sim \\text{Exponential}(\\lambda_i)$$\n", "or\n", - "$$Pr(D_i | \\lambda_i) = \\lambda_i \\text{exp}(-\\lambda_i D_i)$$\n", + "$$f(D_i | \\lambda_i) = \\lambda_i \\text{exp}(-\\lambda_i D_i)$$\n", "It’s the censored cats that are tricky. If something else happened before a cat could be adopted, or it simply hasn’t been adopted yet, then we need the probability of not being adopted, conditional on the observation time so far. One way to motivate this is to image a cohort of 100 cats, all joining the shelter on the same day. \n", "- If half have been adopted after 30 days, then the probability of waiting 30 days and still not being adopted is 0.5. \n", "- If after 60 days, only 25 remain, then the probability of waiting 60 days and not yet being adopted is 0.25. \n", "\n", - "Thus, any given rate of adoption implies a proportion of the cohort of 100 cats that will remain after any given number of days. This probability comes from the cumulative probability distribution. A cumulative distribution gives the proportion of cats adopted before or at a certain number of days. So $1 - \\text{CDF}$ gives the probability a cat is not adopted by the same number of days—also known as the complementary cumulative distribution function (CCDF). If the exponential distribution is used, the CDF is\n", - "$$Pr(D_i | \\lambda_i) = 1 - \\text{exp}(-\\lambda_i D_i)$$\n", + "Thus, any given rate of adoption implies a proportion of the cohort of 100 cats that will remain after any given number of days. This probability comes from the cumulative probability distribution. A cumulative distribution gives the proportion of cats adopted before or at a certain number of days. So $1 - \\text{CDF}$, which is the CCDF, gives the probability a cat is not adopted by the same number of days. Remember from the _Estimating the survival function_ section, this is equivalent to the survival function. If the exponential distribution is used, the CDF is\n", + "\n", + "$$F(D_i | \\lambda_i) = 1 - \\text{exp}(-\\lambda_i D_i)$$\n", + "\n", + "where the complement is (here we use $S$ to denote the equivalence of the survival function and CCDF)\n", "\n", - "where the complement is\n", - "$$Pr(D_i | \\lambda_i) = \\text{exp}(-\\lambda_i D_i)$$\n", + "$$S(D_i|\\lambda) = \\text{exp}(-\\lambda_i D_i)$$\n", "\n", "Which is what we need in our model since it is the probability of waiting $D_i$ days without being adopted yet. The `pm.Censored` from PyMC offers a convenient way to model censored data and the probability density function (PDF) is defined as\n", "\n", @@ -269,14 +227,28 @@ "source": [ "### Implementation in Bambi\n", "\n", - "To understand how this is used, lets use Bambi to recover the parameters of the censored distribution with no predictors. Before the model is fit, `days_to_event` is scaled to represent months as the raw values contain very large values. This scaling ensures a smoother sampling process.\n", + "To understand how this is used, lets use Bambi to recover the parameters of the censored distribution with no predictors. Before the model is fit, `days_to_event` is scaled to represent months as the raw values contain very large values. This scaling ensures a smoother sampling process. \n", "\n", - "The `exponential` distribution is used to model the cat adoption rate parameter. But why not enter `censored` as the likelihood like we normally do in Bambi? The `pm.Censored` is indeed eventually used as the likelihood. However, there also needs to be a distribution that models the rate parameter. In this example it is the `exponential` distribution. This distribution is then used as input into the `pm.Censored` distribution. For more information on how to use the `pm.Censored` distribution, see the following PyMC documentation: [Bayesian regression models with truncated and censored data](https://www.pymc.io/projects/examples/en/latest/generalized_linear_models/GLM-truncated-censored-regression.html#run-the-truncated-and-censored-regressions) and [Censored data models](https://www.pymc.io/projects/examples/en/latest/survival_analysis/censored_data.html)." + "Additionally, modeling censored data in Bambi requires a new formula syntax `censored(time, event)` on the response term. `censored` indicates we want to model censored data and gets parsed where `time` and `event` are passed into a Bambi transformation function [censored](https://github.com/bambinos/bambi/blob/93b2c113333245d9d8b51b4661a218d6a3ce7397/bambi/transformations.py#L12). This function takes two arguments: the first being the observed value $Y$ (in this example `time`), and the second being the type of censoring of the event. In Bambi, it is possible to have `left`, `none`, `right`, and `interval` censoring. `event` needs to be encoded as one of the censoring types. In our cat adoption example, we will encode the adoption event as `right`.\n", + "\n", + "Lastly, the `exponential` distribution is used to model the cat adoption rate parameter. But why not enter `censored` as the likelihood like we normally do in Bambi? The `pm.Censored` is indeed eventually used as the likelihood. However, there also needs to be a distribution that models the rate parameter. In this example it is the `exponential` distribution. This distribution is then used as input into the `pm.Censored` distribution. For more information on how to use the `pm.Censored` distribution, see the following PyMC documentation: [Bayesian regression models with truncated and censored data](https://www.pymc.io/projects/examples/en/latest/generalized_linear_models/GLM-truncated-censored-regression.html#run-the-truncated-and-censored-regressions) and [Censored data models](https://www.pymc.io/projects/examples/en/latest/survival_analysis/censored_data.html)." + ] + }, + { + "cell_type": "code", + "execution_count": 68, + "metadata": {}, + "outputs": [], + "source": [ + "cats = cats_df.copy()\n", + "cats[\"adopt\"] = np.where(cats[\"out_event\"] == \"Adoption\", \"right\", \"none\")\n", + "cats[\"color_id\"] = np.where(cats[\"color\"] == \"Black\", 1, 0)\n", + "cats = cats[[\"days_to_event\", \"adopt\", \"color_id\"]]" ] }, { "cell_type": "code", - "execution_count": 17, + "execution_count": 69, "metadata": {}, "outputs": [ { @@ -323,10 +295,10 @@ "\n" ], "text/plain": [ - "" + "" ] }, - "execution_count": 17, + "execution_count": 69, "metadata": {}, "output_type": "execute_result" } @@ -334,7 +306,7 @@ "source": [ "model_1 = bmb.Model(\n", " \"censored(days_to_event / 31, adopt) ~ 1\", \n", - " data=cats_new,\n", + " data=cats,\n", " family=\"exponential\",\n", " link=\"log\"\n", ")\n", @@ -344,7 +316,7 @@ }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 12, "metadata": {}, "outputs": [ { @@ -390,7 +362,7 @@ "\n", "
\n", " \n", - " 100.00% [4000/4000 00:06<00:00 Sampling 4 chains, 0 divergences]\n", + " 100.00% [4000/4000 00:05<00:00 Sampling 4 chains, 0 divergences]\n", "
\n", " " ], @@ -405,7 +377,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "Sampling 4 chains for 500 tune and 500 draw iterations (2_000 + 2_000 draws total) took 6 seconds.\n" + "Sampling 4 chains for 500 tune and 500 draw iterations (2_000 + 2_000 draws total) took 5 seconds.\n" ] } ], @@ -421,7 +393,7 @@ }, { "cell_type": "code", - "execution_count": 19, + "execution_count": 13, "metadata": {}, "outputs": [ { @@ -441,7 +413,7 @@ }, { "cell_type": "code", - "execution_count": 20, + "execution_count": 14, "metadata": {}, "outputs": [ { @@ -509,7 +481,7 @@ "Intercept 1099.0 1.0 " ] }, - "execution_count": 20, + "execution_count": 14, "metadata": {}, "output_type": "execute_result" } @@ -523,7 +495,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Interpreting the intercept (the cat adoption rate parameter) alone is of not much value. Therefore, lets use the `scipy.stats.sf` survival function to compute the probability of not being adopted after a range of months, given the learned rate parameter $\\lambda$. Additionally, we will plot the average cat adoption waiting time." + "Interpreting the intercept (the cat adoption rate parameter) alone is of not much value. Therefore, lets use the\n", + "survival function to compute the probability of not being adopted after a range of months, given the learned rate parameter $\\lambda$. We could dervive the survival function and pass the intercept parameter to it, but SciPy already implements it as [scipy.stats.expon.sf](https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.expon.html), so we will just use this implementation." ] }, { @@ -537,7 +510,7 @@ }, { "cell_type": "code", - "execution_count": 32, + "execution_count": 15, "metadata": {}, "outputs": [], "source": [ @@ -553,7 +526,7 @@ }, { "cell_type": "code", - "execution_count": 34, + "execution_count": 16, "metadata": {}, "outputs": [ { @@ -568,7 +541,7 @@ } ], "source": [ - "t = np.linspace(0, max(cats_new[\"days_to_event\"] / 31), 100)\n", + "t = np.linspace(0, max(cats[\"days_to_event\"] / 31), 100)\n", "S0 = scipy.stats.expon.sf\n", "cdf = scipy.stats.expon.cdf\n", "\n", @@ -648,14 +621,15 @@ "source": [ "## Implementation in Bambi\n", "\n", - "Adding predictors to model the hazard rate as a function of our predictors is trivial in Bambi. We simply continue to use the formula syntax. In the backend, the rate is modeled as a function of the specified predictors in the Bambi model. For example, if in the Bambi model, we specified `censored(y, event) ~ 1 + x` with a Gaussian likelihood, then the latent rate $\\lambda$ is modeled as a Gaussian distribution with mean and sigma according to\n", + "Adding predictors to model the hazard rate as a function of our predictors is trivial in Bambi. We simply continue to use the formula syntax. In the backend, the rate is modeled as a function of the specified predictors in the Bambi model. For example, if in the Bambi model, we specified `censored(y, event) ~ 1 + x` with an exponential likelihood, then the latent rate $\\lambda$ is modeled as an exponential distribution according to\n", "\n", "$$\\alpha \\sim \\mathcal{N}(0, 1)$$\n", - "$$\\sigma \\sim \\mathcal{HN}(1)$$\n", - "$$\\mu = \\alpha + \\beta X$$\n", - "$$\\lambda \\sim \\mathcal{N}(\\mu, \\sigma)$$\n", + "$$\\beta \\sim \\mathcal{N}(0, 1)$$\n", + "$$\\mu = \\text{exp}(\\alpha + \\beta X)$$\n", + "$$\\lambda = 1 / \\mu$$\n", + "$$Y \\sim \\text{Exponential}(\\lambda)$$\n", "\n", - "where $\\lambda$ is then passed to the `dist` argument of the `pm.Censored` distribution.\n", + "where $Y$ is then passed to the `dist` argument of the `pm.Censored` distribution.\n", "\n", "### Cat adoption rates by color\n", "\n", @@ -664,7 +638,7 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 21, "metadata": {}, "outputs": [ { @@ -724,10 +698,10 @@ "\n" ], "text/plain": [ - "" + "" ] }, - "execution_count": 13, + "execution_count": 21, "metadata": {}, "output_type": "execute_result" } @@ -735,7 +709,7 @@ "source": [ "cat_model = bmb.Model(\n", " \"censored(days_to_event / 31, adopt) ~ 0 + color_id\", \n", - " data=cats_new,\n", + " data=cats,\n", " center_predictors=False,\n", " priors={\"color_id\": bmb.Prior(\"Normal\", mu=0, sigma=1)},\n", " categorical=[\"color_id\"],\n", @@ -748,7 +722,37 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 23, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + " Formula: censored(days_to_event / 31, adopt) ~ 0 + color_id\n", + " Family: exponential\n", + " Link: mu = log\n", + " Observations: 22356\n", + " Priors: \n", + " target = mu\n", + " Common-level effects\n", + " color_id ~ Normal(mu: 0.0, sigma: 1.0)\n", + "------\n", + "* To see a plot of the priors call the .plot_priors() method.\n", + "* To see a summary or plot of the posterior pass the object returned by .fit() to az.summary() or az.plot_trace()" + ] + }, + "execution_count": 23, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "cat_model" + ] + }, + { + "cell_type": "code", + "execution_count": 22, "metadata": {}, "outputs": [ { @@ -794,7 +798,7 @@ "\n", "
\n", " \n", - " 100.00% [4000/4000 05:32<00:00 Sampling 4 chains, 0 divergences]\n", + " 100.00% [4000/4000 06:07<00:00 Sampling 4 chains, 0 divergences]\n", "
\n", " " ], @@ -809,7 +813,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "Sampling 4 chains for 500 tune and 500 draw iterations (2_000 + 2_000 draws total) took 333 seconds.\n" + "Sampling 4 chains for 500 tune and 500 draw iterations (2_000 + 2_000 draws total) took 368 seconds.\n" ] } ], @@ -825,7 +829,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 24, "metadata": {}, "outputs": [ { @@ -907,7 +911,7 @@ "color_id[1] 1652.0 1.0 " ] }, - "execution_count": 15, + "execution_count": 24, "metadata": {}, "output_type": "execute_result" } @@ -934,7 +938,7 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 25, "metadata": {}, "outputs": [ { @@ -953,69 +957,7 @@ }, { "cell_type": "code", - "execution_count": 33, - "metadata": {}, - "outputs": [], - "source": [ - "other_cats = (idata[\"posterior\"][\"color_id\"].sel({\"color_id_dim\": \"0\"})\n", - " .values\n", - " .flatten()\n", - " )\n", - "other_cats_preds = np.quantile(other_cats, [0.025, 0.5, 0.975])\n", - "\n", - "black_cats = (idata[\"posterior\"][\"color_id\"].sel({\"color_id_dim\": \"1\"})\n", - " .values.\n", - " flatten()\n", - " )\n", - "black_cats_preds = np.quantile(black_cats, [0.025, 0.5, 0.975])\n", - "\n", - "lambdas = {\n", - " \"Other cats\": other_cats_preds,\n", - " \"Black cats\": black_cats_preds\n", - " }" - ] - }, - { - "cell_type": "code", - "execution_count": 34, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0MAAAEpCAYAAABRMPn6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACUSklEQVR4nOzdd3xT1fvA8U+SJt0t3S0ttKXsvacs2UsZKoKKoCCKgyHqT3Ew3IJfcICLIUNEprKtyN6j7D1KWaWlLd1t0uT+/ggEStPSYmk6nvfrlVdu7nxyepPTJ/fcc1SKoigIIYQQQgghRBmjtnUAQgghhBBCCGELkgwJIYQQQgghyiRJhoQQQgghhBBlkiRDQgghhBBCiDJJkiEhhBBCCCFEmSTJkBBCCCGEEKJMkmRICCGEEEIIUSZJMiSEEEIIIYQokyQZEkIIIYQQQpRJkgwJIYQQQgghyiRJhoQQQgghhBBlUolNhubMmYNKpbI87OzsCAoKYsiQIVy5cqVQjzV+/HhUKhU3btwo8v3dfp+RkZG5zrO2zo4dOxg/fjw3b94slJjvPbaDgwMXL17Msbxdu3bUrl27wPstaLz3/v1VKhU+Pj60a9eOVatWFfj49+737rK0hW+++QaVSlXgsrx9bj1sD+v8ym/8hf2ZtObe8yu3x6ZNm4rNeZNfJpMJFxcX3nzzTVuHUmZInWWbOqsgFi1aRK1atXB0dESlUnHw4EGr60n9Y11R1D8P8zyR+ufhK651T4lNhm6bPXs2O3fuJDw8nGHDhrFw4UJat25NamqqrUMrFD169GDnzp0EBAQUaJ0dO3YwYcKEh1axZGZm8v777xfa/h403tt//x07dvDTTz+h0Wjo1asXK1eufKA48lPeRWHWrFkAHDt2jN27d9s0Fmse9vlVHOzcuTPbo3v37jg6OuaY37Bhw2Jz3uTXsWPHSE1NpUmTJrYOpcyROss2ddb9xMbG8txzzxEWFsa6devYuXMnVatWzXMbqX+Knq3Pk6JSWuuf4lr32Nk6gP+qdu3aNG7cGID27dtjNBqZNGkSK1as4JlnnrG6TVpaGk5OTkUZ5gPz8fHBx8fnP69T2Lp27cpvv/3G2LFjqVevXpEe+253//1vx+Xh4cHChQvp1atXgfdni7K81759+zh06BA9evRg9erVzJw5k2bNmtk0prKoefPm2V77+PigVqtzzL97eUmxZ88egGJXIZUFUmcVj+/Ze50+fRqDwcCzzz5L27Zt87WN1D/iYSmt9U9xrXtK/JWhe90+UW434bp9OfPAgQM88cQTeHh4EBYWBsC2bdvo0KEDrq6uODk50bJlS1avXp3rvi9dukTfvn1xc3PD3d2dZ599ltjYWMvys2fPMmTIEKpUqYKTkxOBgYH06tWLI0eOPND+IH+Xze9dZ/z48bz11lsAhIaGWi6nTpo0CZVKxcKFC3PsY+7cuahUKvbu3Zvrce729ttv4+XlxTvvvHPfde9XzrnFu2nTpnzFcjcHBwd0Oh1arTbb/DNnzjBw4EB8fX2xt7enRo0afP/99zm2t1aWKpWKY8eOMWDAANzd3fHz8+OFF14gMTExx/Z//vkndevWxd7enkqVKjFt2rQCNx2YOXMmAJ9//jktW7bk999/Jy0tLcd6q1evpn79+tjb2xMaGsrkyZNz3Wd+zvXbcUZEROR5Xt7v75Xfsi5I/Lm532do69athXbO58Xa5/R2eR4+fJgnn3wSd3d3PD09GTNmDFlZWZw6dYquXbvi6upKSEgIX375ZY795rcs8/Lzzz9Tp04dHBwcqF27NuvXr2fPnj3ZvguF7Uid9fDrrPuV2+DBg3nkkUcA6N+/PyqVinbt2t13v/eS+se6wqp/8vO/gtQ/ZsWh/ilRdY9SQs2ePVsBlL1792abP23aNAVQfvrpJ0VRFOWjjz5SACU4OFh55513lPDwcGXFihXKpk2bFK1WqzRq1EhZtGiRsmLFCqVz586KSqVSfv/992z7vHsfb731lrJ+/Xrl66+/VpydnZUGDRooer1eURRF2bx5s/Lmm28qS5YsUTZv3qwsX75c6d27t+Lo6KicPHmywPu7+31euHAh13n3vr506ZLy+uuvK4CybNkyZefOncrOnTuVxMREpUGDBkqrVq1ylGeTJk2UJk2aFKjcb5f1hg0bLMvbtm2r1KpVy/I6P+WcV7z3i2PXrl2KwWBQ9Hq9cunSJeWNN95Q1Gq1sm7dOsu6x44dU9zd3ZU6deooc+fOVf7++2/lzTffVNRqtTJ+/Hir+71dlrf/VtWqVVM+/PBDJTw8XPn6668Ve3t7ZciQIdm2Xbt2raJWq5V27dopy5cvVxYvXqw0a9ZMCQkJUfL7UUtLS1Pc3d0tf4tffvlFAZQ5c+ZkW++ff/5RNBqN8sgjjyjLli1TFi9erDRp0kSpWLFijmPl91zP73mZ198rv2VdkPitKchn6L+e87c9//zzirOzs9Vl1j6nd587kyZNUsLDw5W3335bAZTXXntNqV69uvLNN98o4eHhypAhQxRAWbp0qWX7gpy3uRk5cqTi4OCgvP/++8q///6rTJ8+XfHz81MqVKigdOrUKd/vXfx3UmfZps7KT7mdPXtW+f777xVA+fTTT5WdO3cqx44dy3WfUv/Ypv653/8KUv9cyBGjreqfklb3lPhk6PaXUXJysrJq1SrFx8dHcXV1VaKjoxVFuXNCfPjhh9m2b968ueLr66skJydb5mVlZSm1a9dWgoKCFJPJZJl/ex+jR4/Oto8FCxYogDJ//nyrMWZlZSl6vV6pUqVKtm0Lsr8HqVgURVG++uqrHPPuXjciIsIyb8+ePQqg/Prrr1bfh7Xt9+7dq2RmZiqVKlVSGjdubCmve5Oh/JZzbvHeL457H/b29sr06dOzrdulSxclKCgoR3L12muvKQ4ODkp8fHyO/d5bGX355ZfZth0xYoTi4OCQ7Txp0qSJUqFCBSUzM9MyLzk5WfHy8sp3ZTR37lwFUH744QfL9i4uLkrr1q2zrdesWTOlfPnySnp6umVeUlKS4unpmeNY+f0bFOS8zO3vld+yLkj81jzIZ+hBz/nbHrQymjJlSrZ169evb6nMbzMYDIqPj4/St29fy7yCnLfWLFmyRAFy/KP86aefKoDy3nvvKYqiKCaTSXFxcVFiYmLy3J/4b6TOsk2dld9y27hxowIoixcvvu8+pf6xXf2T1/8KUv9cyBGjLeqf/NY9ilJ86p8S30yuefPmaLVaXF1d6dmzJ/7+/qxduxY/P79s6/Xr188ynZqayu7du3niiSdwcXGxzNdoNDz33HNcvnyZU6dO5TjWve25n3rqKezs7Ni4cSMAWVlZfPrpp9SsWROdToednR06nY4zZ85w4sSJAu/vYRgwYAC+vr7ZLnV+++23+Pj40L9//wLtS6fT8fHHH7Nv3z7++OOPHMsftJwLYu7cuezdu5e9e/eydu1ann/+eV599VW+++47ADIyMtiwYQN9+vTBycmJrKwsy6N79+5kZGSwa9eu+x7nsccey/a6bt26ZGRkEBMTY3mv+/bto3fv3uh0Ost6Li4uBWo7PnPmTBwdHXn66act2z/55JNs3bqVM2fOWI61d+9e+vbti4ODg2VbV1fXHMd6kL/Bg56X+S3rgsR/P/mJtTDP+QfRs2fPbK9r1KiBSqWiW7dulnl2dnZUrlzZ0lSqMM7bSZMm0aRJkxzvsWbNmgCWex3OnDmDh4dHiWlzXtJJnVUw/+Xz+7DrIKl/pP65X6xlsf7Jb90Dxaf+KfHJ0O0vo4iICK5evcrhw4dp1apVjvXu7mUjISEBRVGs9rxRvnx5AOLi4nIs8/f3z/bazs4OLy8vy7pjxozhgw8+oHfv3qxcuZLdu3ezd+9e6tWrR3p6eoH39zDY29szfPhwfvvtN27evElsbCx//PEHQ4cOxd7evsD7e/rpp2nYsCHjxo3DYDBkW/ag5VwQNWrUoHHjxjRu3JiuXbvy448/0rlzZ95++21u3rxJXFwcWVlZfPvtt2i12myP7t27A+Sre0wvL69sr2+X1e2/6+33eu8/NIDVedacPXuWLVu20KNHDxRF4ebNm9y8eZMnnngCuNPDT0JCAiaTKcf5AznPqQf5GzzoeZnfsi5I/PeTn1gL+5wvKE9Pz2yvdTodTk5O2Sri2/MzMjKA/JdlbqKjozl06BADBw7Msezy5cvAnRtYDxw4YKmcYmNjadOmDe+99x6KojzgOxZ5kTqrYP7L5/dh10FS/2Qn9Y/UPwWpe6D41D8lvje5219G93P3DYQeHh6o1WquXbuWY72rV68C4O3tnWNZdHQ0gYGBltdZWVnExcVZvqjmz5/PoEGD+PTTT7Ntd+PGDcqVK1fg/T0sr7zyCp9//jmzZs0iIyODrKwsXn755Qfal0ql4osvvqBTp0789NNP2ZY9aDn/V3Xr1mX9+vWcPn2a2rVrW36FevXVV62uHxoa+p+P6eHhgUql4vr16zmWRUdH52sfs2bNQlEUlixZwpIlS3Is//XXX/n4448tx7K233vnPcjf4EHPSw8Pj3yVtYODQ77jv5/8xlqY53xRyG9Z5uZ2pWPtn5DffvsNf39/goKCAIiIiKBx48YcOnSI/v378/777/Pss88WwrsQ1kidVXAP+vm1RR0k9U/2uKT+KVv1T0HqHig+9U+JT4YehLOzM82aNWPZsmVMnjwZR0dHwDwY1Pz58wkKCrI6vsCCBQto1KiR5fUff/xBVlaWpdcZlUqVI9NfvXo1V65coXLlygXe339x7y9HdwsICODJJ59k+vTp6PV6evXqRcWKFR/4WB07dqRTp05MnDiRChUqWOYXpJzziregbg+U5+Pjg5OTE+3btyciIoK6detma0JQmJydnWncuDErVqxg8uTJluOkpKTkaxA+o9HIr7/+SlhYGL/88kuO5atWrWLKlCmsXbuWnj170rRpU5YtW8ZXX31l+YUnOTk5x/gWD3Ku5+e8tPb3KkhZ5zf++8nvZ6iwz/mH7b+et7ebHBw9ejRbU4UlS5awY8eObE0nIiIiqFSpEn379mX+/Pm0aNGicN6EKDRSZz3Y5/dBy+2/kPone1yFXf/kdp5I/VN4/st5W5C6B4pP/VMmkyGAzz77jE6dOtG+fXvGjh2LTqdj+vTpHD16lIULF1rtinLZsmXY2dnRqVMnjh07xgcffEC9evV46qmnAHPbzDlz5lC9enXq1q3L/v37+eqrr7JlwQXZ339Rp04dAKZNm8bzzz+PVqulWrVquLq6AjBy5EjL2AGzZ8/+z8f74osvaNSoETExMdSqVcsyP7/lfL94c3P06FGysrIA86XdZcuWER4eTp8+fSy/XEybNo1HHnmE1q1b88orrxASEkJycjJnz55l5cqV/Pvvv//5/QNMnDiRHj160KVLF0aOHInRaOSrr77CxcWF+Pj4PLddu3YtV69e5YsvvrD6j0Xt2rX57rvvmDlzJj179mTSpEl07dqVTp068eabb2I0Gvniiy9wdnbOcayCnuv5OS9z+3vlt6wLEn9eCvIZKuxz/mH7L+dtxYoVadKkCf/73//w8fGhbt26bNmyhWnTpgHZmylERERw8OBBJk+eLIlQMSZ11oN9fh+k3PJL6p+ir3/yOk+k/ik8D3reFqTugWJU/9ii14bCkFs3pfe63aNGbGxsjmVbt25VHn30UcXZ2VlxdHRUmjdvrqxcuTLXfezfv1/p1auX4uLiori6uioDBgxQrl+/blkvISFBefHFFxVfX1/FyclJeeSRR5StW7cqbdu2Vdq2bVvg/d39PgvaM4+iKMq7776rlC9fXlGr1QqgbNy4MdvykJAQpUaNGnmW373yKveBAwcqQLbe5BQl/+V8v3itxXH3w93dXalfv77y9ddfKxkZGdnWv3DhgvLCCy8ogYGBilarVXx8fJSWLVsqH3/8sdX93tubz73nT25lvnz5cqVOnTqKTqdTKlasqHz++efKG2+8oXh4eOT6XhRFUXr37q3odLo8e1R5+umnFTs7O0uvU3/99ZdSt27dbMe6He+98vM3KMh5qSi5/73yW9YFif9eBY31tgc552970N587j13ctvPvT0xKkr+y9KaCxcuKF27dlVcXFyUcuXKKb169VJmzpypAMrq1asVRVGUqKgopXz58sq2bduUgIAAJTIy8r77FQ9G6izb1Fn5Lbf/2puc1D9FV//kdZ5I/ZM9RlvUP/mpexSleNU/JTYZEv/NoUOHFED5/vvvbR1KqabX65WaNWsWy37175XXP2GlgZzzOf35559Kly5dFEVRlOnTpysNGjRQ0tLSbByVEDnJ57fgpP4pPuT8zak41T9ltplcWXXu3DkuXrzIe++9R0BAAIMHD7Z1SKXKiy++SKdOnQgICCA6OpoffviBEydOWC4Ri6In53zubrcJB/NNvvv27WPo0KEsWLDAxpEJYSaf3/yT+qf4kfM3d8Wp/pFkqIyZNGkS8+bNo0aNGixevBgnJydbh1SqJCcnM3bsWGJjY9FqtTRs2JA1a9bQsWNHW4dWZsk5n7uPPvoo2+uZM2faKBIhrJPPb/5J/VP8yPmbu+JU/6gURQaTEEIIIYQQQpQ9JX7QVSGEEEIIIYR4EJIMCSGEEEIIIcokSYaEEEIIIYQQZVKJ6EDBZDJx9epVXF1d/9MAaUIIIQpGURSSk5MpX748arX8fnY3qZuEEMI2CrNuKhHJ0NWrV6lQoYKtwxBCiDLr0qVLBAUF2TqMYkXqJiGEsK3CqJtKRDLk6uoKwIULF/D09LRxNMWHwWDg77//pnPnzmi1WluHU6xI2Vgn5ZI7KRvr4uPjCQ0NtXwPizukbrJOPkvWSbnkTsrGOimX3BVm3VQikqHbzQ9cXV1xc3OzcTTFh8FgwMnJCTc3N/mQ3EPKxjopl9xJ2VhnMBgApBmYFVI3WSefJeukXHInZWOdlEvuCrNukgbgQgghhBBCiDJJkiEhhBBCCCFEmSTJkBBCCCGEEKJMKvA9Q1u2bOGrr75i//79XLt2jeXLl9O7d+88t9m8eTNjxozh2LFjlC9fnrfffpuXX375QWMWQhQxk8mEXq+3dRgPlcFgwM7OjoyMDIxGo63DKTJarRaNRmPrMIQQQgibKHAylJqaSr169RgyZAj9+vW77/oXLlyge/fuDBs2jPnz57N9+3ZGjBiBj49PvrYXQtiWXq/nwoULmEwmW4fyUCmKgr+/P5cuXSpznQWUK1cOf3//Mve+hRBCiAInQ926daNbt275Xv+HH36gYsWKTJ06FYAaNWqwb98+Jk+eLMmQEMWcoihcu3YNjUZDhQoVSvWgmyaTiZSUFFxcXEr1+7yboiikpaURExMDQEBAgI0jEkIIIYrWQ+9ae+fOnXTu3DnbvC5dujBz5kwMBoN0FShEMZaVlUVaWhrly5fHycnJ1uE8VLebAjo4OJSZZAjA0dERgJiYGHx9faXJnBBCiDLloSdD0dHR+Pn5ZZvn5+dHVlYWN27csPpLZGZmJpmZmZbXSUlJAOgzMy39ios7faxLmeQkZWNdQcslMzMTRVGws7MrE83kbj+X9vd6LwcHBxRFIT09HXt7+2zL5DN0R251k8FgkHK6i3z/WiflkjspG+uKZbkoJjAawGQAY1b257umVUYDKEYwZd16GG89br3OtsxkmVaZjKDctY1iuvVsvPNaMUJSaqG9pSIZdPXedui3/+nIrX36Z599xoQJE3LM/2vpfPwrViv8AEu48PBwW4dQbEnZWJffcrGzs8Pf35/U1NTi9WX8ECUnJ9s6hCKn1+tJT09n8+bNZGVlZVuWlpZmo6iKn9zqpo0bN5b6K6cPQr5/rZNyyZ2UzT0UBbWSxcZ1f6Ix6VGbDGgUw61nPRqT4a55+ruWGW4t06NWsu48TFnZX99nnkrJQq0YUStGVIrRPA/F1qUCgDaz8OJ46MmQv78/0dHR2ebFxMRgZ2eHl5eX1W3effddxowZY3mdlJREhQoV8Es/Tffuox9qvCWJwWAgPDycTp06SXPDe0jZWFfQcsnIyODSpUu4uLjg4OBQBBHajqIoJCcn4+rqWuY6EsjIyMDR0ZE2bdrk+DvHxcXZKKriJ7e6qX379rnWZ2WRfP9aJ+WSuxJfNiYj6FNBnwKZyaj0yZCZAreeVYZ0MKSCIQ0M6WBIuzUv7a5HOqq7lqM3zy8uyUdeslCThYYsxQ4DGrLQmJ8V87QRDVmoMd1az3j7WTHPv7387vWMt14bFfP0nXnm6ZQsI7CkUOJ/6MlQixYtWLlyZbZ5f//9N40bN871hLe3t8/RVAMgLH4zeiM4O5TAD8pDpNVqS+aXRxGQsrEuv+ViNBpRqVSo1epSfx/N7aZxKpWKuXPnMmrUKG7evGnboIqIWq1GpVJZPS/k83NHbnWTfM9YJ+VinZRL7mxSNopiTmQybkJ6AqTfer77dWbynYf+9nTKnXmGwmuylWuYqDBp7DFp7MlS25Ol0mFQ6dCrdOjRkYmWDEVLumJHmklrfhg1pJrsSDepSTfaYcAO/a2HATv0ijaXeRrLa3NiY0eWorEkOlnY3TWtAaz/gKhSgU6jRmenxt5ObZm2PCyvNeg0t9a5NV9rp0Kn0aDVqNBq1Gg1auw0KvMyjQr71CSYbKNkKCUlhbNnz1peX7hwgYMHD+Lp6UnFihV59913uXLlCnPnzgXg5Zdf5rvvvmPMmDEMGzaMnTt3MnPmTBYuXFjgYINV19m+fROdO3Qq8LZCiLLl0qVLjB8/nrVr11ruT+zduzcffvhhtl/xQ0JCGDVqFG+88YYNo/1v2rVrR/369S29dgohRJmkKOYEJvUGpMZAaiykxEJaXN7JjqmQmoGr7cDe9dbDDUXnglHrjF7tgF5lT4bKgXTFgTRFR6qiI8WkJdGoIzHLjpsGLQkGO+L0dsRm2nEjU01Slo4MzA9DHklHQahU4KjV4KTT4KjT4KS1w1Gnscxz15mfnXR2OGg1OGjV2NuZn3O8ttNgb3mtwd4u+zpajeqhtbSIi4vjlULaV4GToX379tG+fXvL69tNBp5//nnmzJnDtWvXiIqKsiwPDQ1lzZo1jB49mu+//57y5cvzzTffPHC32ikHFoMkQ0KIPJw/f54WLVpQtWpVFi5cSGhoKMeOHeOtt95i7dq17Nq1C09PzyKPS3rQFEKIAjIZISUme3KTGnvr9Y1by2LvPExZ99+nNWotOHqAYznzs0O5O6/t3cDeFaPWmVScSFYcSDA6kJBlzw2Dlli9PdczNFxPg/g0A/GpeuLj9SSk6TEY/1szNxUKzvZavOztcHGww9neDld7O5ztNbjYa3Gx1+B8a5mLvflxZx27O0mPzjxtb6cuc03B76fAyVC7du0sHSBYM2fOnBzz2rZty4EDBwp6KKsaJG/ifEwylXxdC2V/QojS59VXX0Wn0/H3339buo6uWLEiDRo0ICwsjHHjxjFjxgzatWvHxYsXGT16NKNHm+9HNBqNlv2sX7+eUaNGcenSJR555BFmz56drQfM2bNn8+WXX3LhwgVCQkJ44403GDFiBACRkZGEhoayaNEipk+fzq5du5gxYwZDhgzJEe/Nmzd5++23+fPPP0lMTKRy5cp8/vnn9OzZk7i4OF577TW2bt1KfHw8YWFhvPfeewwYMACAwYMHs3nzZjZv3sy0adMA8xV7d3d3XnvtNf7++29SUlIICgrivffes3p8IYSwCUWB1FhUcZEE3NyHeu9lSLkGiVcg6Yr5OfmaufewgrB3B2dvcPE1Pzt555LoeKA4uJOiduV6uoaY5ExikjO5npTB9aRMridnEBuTyY3UTOJT9SSmG7jzL7AJSL/1yJvOTo27oxZ3Ry1uDnZ3pm893552c7j7tR1Odio2b/ibnj26yA9pD1GR9CZXWPRoCVVfZ87Wf6nU73FbhyNEmaMoCumGAlZKhcRRq8nXr1nx8fGsX7+eTz75xJII3ebv788zzzxjSVCWLVtGvXr1eOmll3jxxRez9SSXlpbG5MmTmTdvHmq1mmeffZaxY8eyYMECAH7++Wc++ugjvvvuOxo0aEBERATDhg3D2dmZ559/3rKfd955hylTpjB79myr95uYTCa6detGcnIy8+fPJywsjOPHj1vG+8nIyKBRo0a88847uLm5sXr1ap577jkqVapEs2bNmDZtGqdPn6Z27dpMnDgRAB8fH0aOHMnx48dZu3Yt3t7enD17lvT0+1faQghRaIwGuBkFcecgMSp7kpN0GZKuglGPHdAU4EIu+1Gpwdkn++N2ouPse+v1rflO3qA1dwSjzzJxLTGdqzcziEnOICYpk+tXM7h+K+GJTb7B9aTLpOnzX6+pVFDOUYuHsw4vZx2ezjo8ne3xdNbi6Wx/17w7Dwftg43fZjAYUMtFnIeuRCVDcX6P4H1zExxbgbHPY2jkDBGiSKUbjNT8cL1Njn18YhecdPf/yjpz5gyKolCjRg2ry2vUqEFCQgKxsbGWQUZdXV3x9/fP1j2ywWDghx9+ICwsDIDXXnvNkmwATJo0iSlTptC3b1/A3CT4+PHj/Pjjj9mSoVGjRlnWseaff/5hz549nDhxgqpVqwJQqVIly/LAwEDGjh1ref3666+zbt06Fi9eTLNmzXB3d0en0+Hk5IS/v79lvaioKBo0aEDjxo0B871RQghR6ExGSLwM8efMSU/cuTvTNy/mo9maCsXFl5smZ9wr1kJdrgK4BYJ7ILgFmZ9d/ECdM6EwGE1cu5nB5YQ0Ll9J51JCKpcTbphfJ6QTnZRBHo2ZsnF1sMPX1R4/Nwf83BzwdbPHz9X87O1ib0lsyjlqsdOU7g6FypoSlQx5NOwD/26iXdZ2tpyOoX11v/tvJIQQd7nfOGe3OTk5WRIhgICAAGJiYgCIjY3l0qVLvPjiiwwbNsyyTlZWFu7u7tn2czsZyc3BgwcJCgqyJEL3MhqNfP755yxatIgrV65YBv50dnbOc7+vvPIK/fr148CBA3Tu3JnevXvTsmXLPLcRQgirFMXcXC3uHMSdvZXsnDdPJ1wAoz73be0cwbMSeATnTHLcAsE1gCxFxZY1a+jevTvqu5qDKYpCdFIGF84ncDkh3ZLk3J6OTsrAdJ9kx95OTflyjvi53ZXouNrj6+aA363kx9fNPl8/tonSqUT95TVVHsWw0Z4Q9XV+276R9tWftnVIQpQpjloNxyd2sdmx86Ny5cqoVCqOHz9O7969cyw/efIkHh4eeHt757mfe9tnq1QqSyJ1uxvun3/+mWbNmmVb73bzttvul7Tc25TvXlOmTOF///sfU6dOpU6dOjg7OzNq1Cj0+jz++QC6devGxYsXWb16Nf/88w8dOnTg1VdfZfLkyXluJ4Qo47IyIfYkRB+59Thqfs5MzH0bjQ48QsErzJz4eFW+NR0GrgFwn6EZMtMyuJoKa49GczE+g3OxKZyLTeV8bAqp92nCZm+nJsjDkSAPp3uezdPeLjrpMEDkqUQlQ+icSQ/piPbCajwjVxOf2hdPZ52toxKizFCpVMX+1zMvLy86derE9OnTGT16dLZkIzo6mgULFjBo0CBL5ajT6bJ1mpAffn5+BAYGcv78eZ555pn/FG/dunW5fPkyp0+ftnp1aOvWrTz++OM8++yzgDkRO3PmTLZmgLm9Bx8fHwYPHszgwYNp3bo1b731liRDQog70uLh+tG7Ep8j5kTIWtM2lcZ8dccz7K5kp5L52b2C1WZsd1MUhbhUPedizInOudgUzt9Kei4lpKEodnD4cI7tNGoVwZ5OVPCUZEc8HMX7vxor3Bo9ARdW0021iz8jLjPkkUr330gIUaZ89913tGzZki5duvDxxx9n61o7MDCQTz75xLJuSEgIW7Zs4amnnkKv1+Pm5pavY4wfP5433ngDNzc3unXrRmZmJvv27SMhIcEy5EB+tG3bljZt2tCvXz++/vprKleuzMmTJ1GpVHTt2pXKlSuzdOlSduzYgYeHB19//TXR0dHZkqGQkBB2795NZGQkLi4ueHp6Mn78eBo1akStWrXIzMxk1apVud5HJYQo5RTFfP/OtcN3kp7rRyHxkvX1HcqBfx3wr3vruQ54VwW7/P0Arc8ycfp6MsevJnH0aiLHryZxJiaFxPTcx/Nx1ChUK1+Oyr6uVPJxJszHhTAfFyp6OqGzk3t0xMNT4pIhqnYhS+1AMDHs371ZkiEhRA5VqlRh3759jB8/nv79+xMXF4e/vz+9e/fmo48+yjbG0MSJExk+fDhVqlQhMzMz31eJhg4dipOTE1999RVvv/02zs7O1KlTh1GjRhU43qVLlzJ27FgGDBhAamqqpWttgA8++IALFy7QpUsXnJyceOmll+jduzeJiXearIwdO5bnn3+emjVrkp6ezoULF9DpdLz77rtERkbi6OhI69at+f333wscmxCiBDKZIOY4RO2Ei9vh4g5IuW59XY8Q8KudPfFxDzJ3m5YPafosTlxL5tjVRI5dMSc/p68nWx1fR6WCwHKOlkQnzNec9AR72LN78wZ69GgmXUiLIlfykiGdM6YqneDUSmolbODolb7UDnS//3ZCiDIlODiY2bNn33e95s2bc+jQIUwmE0lJSQCWpmV36927d44x1gYOHMjAgQOt7jckJCTPMdnu5unpyaxZs3JdtmLFijy3r1q1Kjt37sw27/333+f999/P1/GFECWc0QDXDt1KfHaak6CMm9nXUWvBr2b2Kz5+tcAh//9DJaYZzEnPrSs+x64mcT42xWonBm4OdtQq706t8m7UCnSjur8bod7OVruZNhgM+c29hCh0JS8ZAnR1+8GplXRX72bW3ihqB9axdUhCCCGEEEXDkA6X95mv+FzcDpf3giEt+zpaZ6jQFIJbQXBLCGxkGX8nP0wmhdMxyey9EM/uC/EcvHSTywnWxyrzdbU3Jz3l3akdaH4O8nCUe3lEiVAikyGqdMaocSSYGE4d3EZGj5oPPKCVEEIIIUSxlpUJkdsgcqs5AbpyAEz33H/j6AEVW5gTn+CW5qs/mvw3OTMYTRy9ksjeyHj2XIhnb2SC1Xt8Kng6Uttyxcf87Oua/yRLiOKmZCZDOmdU1brA8RW0zdrOPycep2fd8raOSgghhBCicKTfhLP/wMlVcOYf0CdnX+7iDyGtbiVArcCn+n27sL5bhsFIRNRN9lyIZ09kHAcu3iTdkP2eSSedhoYVPWga6knjYA9qlXfH3Unu6RGlS8lMhgB1rT5wfAU91Lv4YO8lSYaEEEIIUbIlXoZTa80JUOS27F1cu/hB5Y53rvx4hOa7kwOApAwD+yMT2H0hnr2R8Ry+fDNHJwflnLQ0DvakWagnTUI9qVXeDa1GenITpVuJTYao0hmTnSMVs2JJOLeHqzfrUr5c3oMXCiGEEEIUG4oC14/BqTXmBOjaoezLvatB9e5QrYf5np8CXPlRFIUzMSlsOBHDhhPXORCVkKOjAz83e5qGetE01JOmIZ5U8XVBrZb7fETZUnKTIZ0T6mpd4dhyuqt3s+xAV157tIqtoxJCCCGEyJ0pCy7supMA3Yy6a6EKKjS7kwB5Vy7QrjOzjOy5EG9OgE5e51J89g4PQr2daRLiYU6AQjyp4CmdHAhRcpMhgFp94Nhyeqh38ey+S7zavrJ8qIUQQghRvJiMqM5toMHFH7GbOhLSE+4ss3OASu3NCVDVruDiW6Bd30jJZOPJGP49GcOW07Gk6u/c96OzU9MyzIsONfx4tLovgdKCRogcSnYyVLkTitaJCoZY3BKOsedCPZpV8rJ1VEIIIYQQkHQNIubDgbnYJUZR8fZ8R09z4lO9B4S1B51zvnepKAono5PZcOI6G07GcPDSTe4e0szH1Z4O1X15tLovj1TxxklXsv/VE+JhK9mfEJ0Tqqpd4dgyemh2sXh/G0mGhBBCCGE7JiOc3QD758DpdaCYr9QoDuW44NKYit3ewC6kFWjy/y+Y0aSw/ewNwo9f59+TMVy5mb35W+1ANzpU96NDDV9ql3eX+36EKICSnQwB1OoNx5bRU7OLToevMv6xWrjYl/y3JYR4uCIjIwkNDSUiIoL69esXyj5VKhXLly+nd+/ehbI/IUQJknjFchWIpMt35ldsCY0Gk1WlG0fCN1KhYst8J0KnryezdP9llkdcISY50zLfQavmkcrePFrd3PzN313G+RHiQZX8rKFyJxStM0GGG1TJOsPqw7Xo36Ti/bcTQpRagwcP5tdff7W89vT0pEmTJnz55ZfUrVvXhpEVrk2bNtG+fXsSEhIoV66crcMRouwxGeFMuPkq0Jn1oJjM8x09oN5AaPQ8+FQzzzPkHMDUmvhUPX8dvMLSA1c4ciXRMr+ck5ZutQPoVNOXFpW8cdTJYPNCFIaSnwzpnFBV6wpHl5qbyu1rLMmQEIKuXbsye/ZsAKKjo3n//ffp2bMnUVFR99lSCCHuI/EyHJgHEfMg6cqd+cGtoNEQqNELtPm/WqPPMrHpVAxLD1zm35MxlvF/7NQq2lf3pV/DIB6t7ovOTsb8EaKwlY5PVc3eAPTQ7GbfxXjOxabYNh4hhM3Z29vj7++Pv78/9evX55133uHSpUvExsZaXd9oNPL6668TFhaGo6Mj1apVY9q0aTnWmzVrFrVq1cLe3p6AgABee+21XGOYOHEifn5+HDx4MNd1/vrrLxo3boyDgwPe3t707dvXsmz+/Pk0btwYV1dX/P39GThwIDExMYC5mV/79u0B8PDwQKVSMXjwYACWLFlCnTp1cHR0xMvLi44dO5Kamnq/IhNC5MVkMg+IuuApmFoHNn9uToQcPaHFa/DqXhiyBuo+ma9ESFEUjl5JZPxfx2j+2QZemref9ceuYzAq1A5046NeNdn9Xgd+HtSYrrX9JRES4iEp+VeGAKp0gltN5eqpzrFkf2Xe6Vrd1lEJUfooChjSbHNsrVOBRlu/W0pKCgsWLKBy5cp4eVnvZMVkMlG+fHl+//13fH192bFjBy+99BIBAQE89dRTAMyYMYMxY8bw+eef061bNxITE9m+fXuOfSmKwqhRo1ixYgXbtm2jShXrY6CtXr2avn37Mm7cOObNm4der2f16tWW5Xq9nkmTJlGtWjViYmIYPXo0gwcPZs2aNVSoUIGlS5fSr18/Tp06hZubG46Ojly7do0BAwbw5Zdf0qdPH5KTk9m6dSuKoliNQQhxH4piHhNo46dw/eid+SGtodFgqN6zQFeBYpIzWX00iqX7r3DqerJlvo+rPX0aBNK3YSDV/d0K8Q0IIfJSOpIhrSPcairXXbObmftr8Wanqthp5FcUIQqVIQ0+LW+bY793tUDdz65atQoXFxcAUlNTCQgIYNWqVahzGcFdq9Xy7rvv4ubmhlqtJjQ0lB07dvDHH39YkqGPP/6YN998k5EjR1q2a9KkSbb9ZGVlMWjQIPbt28f27dsJCgrKNcZPPvmEp59+mgkTJljm1atXzzL9wgsvWKYrVarEN998Q9OmTUlJScHFxQVPT08AfH19LfcMnTt3jqysLPr27UtwcDAAderUuW95CSHuoSjmXuE2fgxXI8zz7N2g4SBzEuSd/4HeTSaFDSdi+OGEmlO7NmO69duEzk5N55p+9GsUROvK3vJ/ixA2UDqSITAPwHp0KY/Z7eaz5IFsORPLo9X9bB2VEMJG2rdvz4wZMwCIj49n+vTpdOvWjT179liShHvNmjWL3377jYsXL5Keno5er7f0NBcTE8PVq1fp0KFDnscdPXo09vb27Nq1C29v7zzXPXjwIMOGDct1eUREBOPHj+fgwYPEx8djMplvzo6KiqJmzZpWt6lXrx4dOnSgTp06dOnShc6dO/PEE0/g4eGRZyxCiLuc3wwbP4FLu82vtc7Q/GVzczgnz3zvJjPLyJ8RV/lxyznOxaZy++6ERsEe9GsYRI+6Abg7ah/CGxBC5FfpSYYqdwSdCwH6G9RXnWPxvgBJhoQobFon8xUaWx27AJydnalcubLldaNGjXB3d+fnn3/m448/zrH+H3/8wbhx45g8eTItW7bE1dWVr776it27zf8MOTrmb+T2Tp06sXDhQtavX88zzzyT57p57TM1NZXOnTvTuXNn5s+fj4+PD1FRUXTp0gW9Xp/rdhqNhvDwcHbs2MHff//Nt99+y7hx49i9ezehoaH5eg9ClFkXd5qToMit5td2DtBkKDwyGpzz/nHjbkkZBn7bHcWsbRcsXWK7OtjR1FPPO0+2pmpAuYcQvBDiQZSeZEjraB7N+egSumt289WJKsSn6vF01tk6MiFKD5WqQE3VihOVSoVarSY9Pd3q8m3bttG0aVNeeeUVS1O6c+fOWZa7uroSEhLChg0bLB0XWPPYY4/Rq1cvBg4ciEaj4emnn8513bp167JhwwaGDBmSY9nJkye5ceMGn3/+ORUqVABg37592dbR6czfb0ajMcd7bdWqFa1ateLDDz8kODiY5cuXM2bMmFxjEaJMu7If/v0Ezm0wv9bozE3hWr8Jrv753k10Ygazt19gwe4oUjKzAPB3c+DFR0Lp1yCArf/+Tah3yfwOFaK0Kj3JENxqKreE3ro9fJo2kBURV3jhEfklVIiyKDMzk+joaAASEhL47rvvSElJoVevXlbXr1y5MnPnzmX9+vWEhYUxb9489u7dm+1qyvjx43n55Zfx9fWlW7duJCcns337dl5//fVs++rTpw/z5s3jueeew87OjieeeMLqMT/66CM6dOhAWFgYTz/9NFlZWaxdu5a3336bihUrotPp+Pbbb3n55Zc5evQokyZNyrZ9cHAwKpWKVatW0b17dxwdHTl27BgbNmygc+fO+Pr6snv3bmJjY6lRo8Z/KU4hSqfoI+aOEU6tMb9W20GDZ6H1WChXId+7OXM9mZ+2nGfFwSuWbrGr+LowvG0Yj9Urj85OjSGf4wwJIYpW6UqGKncAnQu++ljqq87xxz43hrQKQfWAPVAJIUqudevWERAQAJiv6lSvXp3FixfTrl07q+sPHz6cvXv3MmDAAFQqFQMGDGDEiBGsXbvWss7zzz9PRkYG//vf/xg7dize3t65JjpPPPEEJpOJ5557DrVana3L7NvatWvH4sWLmTRpEp9//jlubm60adMGAB8fH+bMmcN7773HN998Q8OGDZk8eTKPPfaYZfvAwEAmTJjA//3f/zFkyBAGDRrEO++8w5YtW5g6dSpJSUkEBwczZcoUunXr9qBFKUTpE3MSNn0Kx/80v1apod4AaPMWeOb/R9S9kfH8uPkc/5yIscxrGurJy20r0a6qL2q1/P8hRHFXupIhrSNU6wZHFvOYdjcToytz9EoSdYLcbR2ZEKIIzZkzhzlz5uS5TkhISLbupu3t7fn++++ZN29eth7nPvvss2zbDR8+nOHDh1vd573dVz/11FOWnuhy07dvX6uJEsCAAQMYMGBAnsf44IMP+OCDD7LNW7duXZ7HFKLMijsHmz6HI4sBBVBB7X7Q7v/y3TucyaTwz4nr/LjlPPsvJgDmFsRdavrzUttKNKwonZUIUZKUrmQIzAOwHllMb91eJuoHsnj/JUmGhBBCiLLMmAU7ppkTIeOtDkhq9IJ274Gf9Z4Z72Uwmlh24DI/bjnP+VjzIMY6jZp+jQIZ2roSYT4uDyt6IcRDVPqSoVu9ynnqY2igOsuKCC3vda+Bg1Zj68iEEEIIUdSuH4MVI+DaQfPrsEeh43gIqJfXVtlsOhXDx6tPcDYmBQA3BzuebR7M4FYh+Lrmf8BVIUTx80Cje02fPp3Q0FAcHBxo1KgRW7duzXP9BQsWUK9ePZycnAgICGDIkCHExcU9UMD3pXWAat0B6O+0j6SMLMKPX384xxJCCCFE8WQ0wKYv4Me25kTIwR16/wDPLst3InQ2JoUhs/cwePZezsak4OmsY1z3Gux4twNvd60uiZAQpUCBk6FFixYxatQoxo0bR0REBK1bt6Zbt25ERUVZXX/btm0MGjSIF198kWPHjrF48WL27t3L0KFD/3PwuarVG4Dumj2oMPHHvksP71hCCCGEKF6uHYKf2ps7STAZoFoPeHUP1B9gvsHnPhLTDExYeYyuU7ew8VQsdmoVQx8JZePYdgxrUwkX+9LXsEaIsqrAn+avv/6aF1980ZLMTJ06lfXr1zNjxowcNxoD7Nq1i5CQEN544w0AQkNDGT58OF9++eV/DD0PYR1A54qb/joNVGfZdlbNlZvpBJbL36CJQgghhCiBsjJh85ew7X+gGMHRE7p/Ze4kIR9JUJbRxMI9UXwdfpqENHNX2B1r+PJe9xpUknuChCiVCnRlSK/Xs3//fjp37pxtfufOndmxY4fVbVq2bMnly5dZs2YNiqJw/fp1lixZQo8ePR486vvROph7lQNe9DyEosCivXJ1SIgHdW8PZqJ0MZlMtg5BiP/uyn5zk7itk82JUM3e5qtBdZ7IVyK09Uws3b/Zygd/HiMhzUBVPxfmvdiUX55vIomQEKVYga4M3bhxA6PRiJ+fX7b5fn5+lsEN79WyZUsWLFhA//79ycjIICsri8cee4xvv/021+NkZmaSmZlpeZ2UlASAwWDI96Blquq9sDvyB48ad6CiH3N3RPJCiwo4l6JL27fLQgZyy0nKxroHLZfY2Fi8vLxK9ZhdiqKg1+tJT08v1e/zboqiYDAYiI2NRaVSoVKpcpwb8hm6ozDqprKgyL9/szJQb/kC9a7vUSkmFGcfjF2/RKne63ZAeW4eGZfKZ2tP8++pWAA8nLSMfDSM/o2DsNMU3mCpUi/lTsrGOimX3BVmmaiUAvzke/XqVQIDA9mxYwctWrSwzP/kk0+YN28eJ0+ezLHN8ePH6dixI6NHj6ZLly5cu3aNt956iyZNmjBz5kyrxxk/fjwTJkzIMf+3337DyckpX7GqTXq6HnkdrSmdIYxnY0ZVegcbaV9efuEWoiB0Oh2enp7Y2ZWeHxLEHYqikJaWRmJiotUrRGlpaQwcOJDExETc3NxsEGHxURh1kyhcHilnaBD1C66Z1wC45NGCo0HPordzve+2aVnw92U1W6JVGBUVapVCaz+FrhVMOMnXnRDFWmHWTQVKhvR6PU5OTixevJg+ffpY5o8cOZKDBw+yefPmHNs899xzZGRksHjxYsu8bdu20bp1a65evWoZIf5u1n59q1ChAteuXcPLyyvfb07z5yuojy7mRPAzdDvVAz9XezaMaY293QN1olfsGAwGwsPD6dSpE1qt1tbhFCtSNtY9aLkYjUaysrJKdXO5rKwsduzYQcuWLctM4qdSqdBoNGg0mlyvhsXFxREQECDJEIVXN5V2RfL9a0hDvelT1Ht+RIWC4uKHsdtklKrd7rup0aTwx/7LTN1wlvhU86/Lbat6827XaoT5OD+ceJF6KS9SNtZJueSuMOumAtX4Op2ORo0aER4eni0ZCg8P5/HHH7e6TVpaWo5/LDQa85g/uf1jZW9vj729fY75Wq22YCdDnX5wdDHV4//F37U30cmZrDpynaebVsz/PkqAApdLGSJlY11By6UslKHBYCArKwsXF5cy8X7zS8rijkKrm8qIh1Yukdvhr9cg/rz5db2BqLp+ip2jx3033XU+jvF/HeNkdDIAYT7OfNCzJu2q+RZ+nLmQ8yV3UjbWSbnkVJjlUeCfP8eMGcNzzz1H48aNadGiBT/99BNRUVG8/PLLALz77rtcuXKFuXPnAtCrVy+GDRvGjBkzLM3kRo0aRdOmTSlfvnyhvRGrwh4FezdUydd4r1EKb2y354fN53iycQU06rJxT4AQQghRKphMsPFj2DrF/Nq1PPSaBlU7570dkJll5Kt1p/hl2wUA3B21jOpYhWebB6PVlI7WIkKIB1PgZKh///7ExcUxceJErl27Ru3atVmzZg3BwcEAXLt2LduYQ4MHDyY5OZnvvvuON998k3LlyvHoo4/yxRdfFN67yI2dvXkA1sO/0824iXJOPYmMS2PNkWv0qveQEzEhhBBCFA59GiwfDif+Mr9uOAg6f2weSPU+zlxP5o3fD3LimrnDiwFNK/B2l+p4OOseZsRCiBLigRrGjxgxghEjRlhdNmfOnBzzXn/9dV5//fUHOdR/13AQHP4d7dE/GN50EF9sus70TefoWTegzPQYJYQQQpRYydGw8Gm4GgEaHTz2LdR7+r6bKYrC/F0X+Xj1CTKzTHg66/iyX1061vS777ZCiLKj9F8bDm4JfnUgK53BDltx0mk4cS2JTadjbR2ZEEIIIfISfQR+7mBOhBw9YdCf+UqEbqRkMvTXfXzw5zEys0y0qerDulGtJRESQuRQ+pMhlQqaDQfAMWIWzzQJBGDGxnO2jEoIIYQQeTm9HmZ1haTL4FUFhm0w/8B5H5tOxdB16lY2nIxBp1HzYc+azBncBF9XhyIIWghR0pT+ZAjMo087ekJiFCPKn0GrUbEnMp69kfG2jkwIIYQQd1MU2DXD3DROnwKhbWBoOHhWynOzDIORCSuPMXj2Xm6kZFLVz4U/X2vFC4+EopZOk4QQuSgbyZDWERoNBsDjyCyeaBQEwPSNZ20YlBBCCCGyMWbBmrGw7v9AMUHD5+HZZXCfbrNPRSfT+/vtzN4eCcDgliH89doj1Ago22NjCSHur2wkQwBNXgSVBiK38lpNPWoVbDwVy/GrSbaOTAghhBAZifDbU7D3F0Bl7i2u1zTQ5D6eiKIozNl+gV7fbeNkdDLeLjpmD27C+Mdq4aDVFF3sQogSq+wkQ+5BUKMXAIGnf6V7nQAAZmyWe4eEEEIIm0q4CDM7w7kNoHWC/vOh5evm+35zEZucyZA5exm/8jj6LBPtq/mwblQb2lcvugFUhRAlX9lJhgCav2J+PvwHrzX3BGD14atE3ki1YVBCCCFEGXZpD/zSAWJPgmsADFkLNXrmucm/J6/TdeoWNp2Kxd5OzcTHazFrcBO8XeyLKGghRGlRtpKhCs0goB5kZVD96nLaV/PBpMCPW+TqkBBCCFHkjiyBOT0hNRb868Kwf6F8/VxXzzAY+fDPo7wwZx9xqXqq+7uy8vVHGNQiRMYOFEI8kLKVDKlU0Oxl8/SeXxjRNgSApfuvcD0pw3ZxCSGEEGWJosCmL2Dpi2DMhGo9zFeE3MrnusmNlEz6/7SLuTsvAvDiI6GseLUVVf1ciypqIUQpVLaSIYBafcHJG5Iu0yRjJ01CPNAbTfyy9bytIxNCCCFKv6xMWD4cNn1qft3ydeg/D+xdct0k8kYq/Wbs4NClm5Rz0vLrC035oGdN6SRBCPGflb1kSOsAjYeYp3f/yIh2lQFYsDuKm2l6GwYmhBBClHLpN2Hu43B4EajtzL3Fdf4Y1LknNRFRCfSdsYOLcWlU8HRk6SstaVvVp+hiFkKUamUvGQJo/IL5S/jidtq5X6NGgBtpeiNzdkTaOjIhhBCidDKkw8IBELUT7N3h2aWWMQBz88/x6wz4eRfxqXrqBLqz7JVWhPnkfgVJCCEKqmwmQ27loebjAKj2/MSIdmEAzNkRSWpmli0jE0IIIUofYxYsHgJRO8yJ0JDVUKldnpss2H2Rl+btI8Ngol01H35/qTk+rtJbnBCicJXNZAjudKRweDHdw3SEeDlxM83Awj1Rto1LCCGEKE0UBVa+AafXgp0DDPwd/OvksbrC5PWnGLf8KCYFnmocxM+DGuNsb1eEQQshyoqymwwFNYHyDcCYiebAHIa3NV8d+mXrBTKzjDYOTgghhCglwj+EgwtApYEn50Bwy1xX1WeZeHPxIb7beBaAUR2r8EW/umg1ZfffFSHEw1V2v13u7mZ770z61vPFz82e6KQMlh+4YtvYhBBCiNJg+zTY8Y15+rFvoVq3XFdNzjDw4q97WXbgChq1ii/61WFUx6oyfpAQ4qEqu8kQQK0+4OwDyVexP7OGYa0rAfDjlvMYTYqNgxNCCCFKLtWh38xXhQA6TYIGz+S67vWkDPr/uIutZ27gqNXwy6DG9G9SsYgiFUKUZWU7GbKzN/csB7D7RwY0rUg5Jy0XbqSy9ug128YmhBBClFD+iQfQrB5tftHyDWj1Rq7rno1Jpu/0HRy/loS3i45Fw5vTvrpvEUUqhCjrynYyBLe62dbCpV04xx3h+RYhAEzfeA5FkatDQgghREGoonbQ+ML3qBQj1H8GOk3Mdd09F+LpN2MnV26mE+rtzLJXWlE3qFzRBSuEKPMkGXL1NzeXA9j9E4NbhuCk03D8WhKbTsfaNjYhhBCiJIk+guaPZ9EoBkxVukKvb8z36Fqx5sg1np25m8R0Aw0rlmPpKy2p6OVUxAELIco6SYbgTkcKR5fgoSQysKm5nfKMjedsGJQQQghRgsRfgPn9UGUmccO5GsY+P4PGenfYM7dd4NXfDqDPMtG5ph8LhjbH01lXxAELIYQkQ2ZBjSCwMRj1sH8OQ1tXQqtRsScynn2R8baOTgghhCjekq/DvD6Qch3Ftxa7K40CrWOO1UwmhY9XHWfSquMoCjzXPJgZzzbCUacp+piFEAJJhu6wdLP9C/4uGvo1DAJg+ia5OiSEEELkKiMR5veDhAtQLpispxeRZeecYzVFUfh0zQl+2XYBgHe6Vmfi47XQqKXrbCGE7UgydFvNx8HFD1Ki4fifDG8bhloF/56M4fjVJFtHJ4QQQhQ/hgxYOACuHwFnX3huufleXCt+2nLekgh92a8ur7QLkzGEhBA2J8nQbXY6aPyieXr3D4R6O9O9TgAAMzbL1SEhhBAiG2MWLHkBLm4Hezd4dgl4hVlddcn+y3y29iQA73WvzlNNKhRlpEIIkStJhu7WeIi5m+3Le+Hyfl5pZ/5SX334KqevJ9s4OCGEEKKYUBRYNRJOrQaNPTz9GwTUs7rqhhPXeWfpYQBealOJl9pYT5iEEMIWJBm6m4sv1O5nnt7zI7XKu9Ollh8mBSasPCbjDgkhhBAAGyZAxHxQqeGJWRDa2upq+y/G8+pvBzCaFPo2COT/ulYv4kCFECJvkgzdq9lL5uejyyD5OuO610Rnp2b72TjWHY22bWxCCCGEre2cDtv+Z57uNQ1q9LS62pnrKbwwZx8ZBhPtq/nwxRN1UUtnCUKIYkaSoXsFNoKgpmAywP7ZVPRyYnibSgB8vPoE6XqjjQMUQgghbCRqN/w9zjzd4SNoOMjqavGZMGTufhLTDTSoWI7vn2mIViP/cgghih/5ZrKm2XDz896ZkKVnRLvKlHd34MrNdH6QzhSEEEKURRlJsGwoKCao2x8eGW11tYQ0PT+c0HA9KZPKvi7Mer4JTjrrg68KIYStSTJkTc3HwTUAUmPg+AocdRrG9agJwA+bz3EpPs3GAQohhBBFbM1YuBkF5YKh+2Sw0i12mj6LYfMiuJ6uwt/NnrkvNMXDWWeDYIUQIn8eKBmaPn06oaGhODg40KhRI7Zu3Zrn+pmZmYwbN47g4GDs7e0JCwtj1qxZDxRwkdBoocmtbrZ3zQBFoXsdf1pU8iIzy8THq4/bNj4hhBCiKB1eDIcXmTtM6PszOLjlWMVgNDFiwQEOXU7EyU5h1vONKF/O0QbBCiFE/hU4GVq0aBGjRo1i3LhxRERE0Lp1a7p160ZUVFSu2zz11FNs2LCBmTNncurUKRYuXEj16sW8R5lGQ8zdhV49AJf3oVKpGP+YeaTs9ceus/VMrK0jFEIIIR6+hIuweox5us3bULFZjlVMJoW3lxxm06lYHLRqXqpupIqvSxEHKoQQBVfgZOjrr7/mxRdfZOjQodSoUYOpU6dSoUIFZsyYYXX9devWsXnzZtasWUPHjh0JCQmhadOmtGzZ8j8H/1A5e0OdJ8zTu38AoJq/K881DwZg/F/HMBhNtopOCCGEePhMRlg+HDKTzJ0LtXnL6mqfrT3B8ograNQqvn26HqGuRRynEEI8oAIlQ3q9nv3799O5c+ds8zt37syOHTusbvPXX3/RuHFjvvzySwIDA6latSpjx44lPT39waMuKk1vdbN9fIW5nTQwulNVvJx1nItN5dcdkTYLTQghhHjotn4NUTtB5wp9fwJNzo4Qftx8jp+3XgDgy351aVfVp6ijFEKIB1ag7l1u3LiB0WjEz88v23w/Pz+io62PwXP+/Hm2bduGg4MDy5cv58aNG4wYMYL4+Phc7xvKzMwkMzPT8jopKQkAg8GAwWAoSMj/jU8tNCFtUEduwfTPBIyP/4CTHbzZqTLvrTjO//45Tfdavvi42hddTHe5XRZFWiYlhJSNdVIuuZOysU7K445iUzcVEdWV/Wg2fYYKyOr6BYprENzzPpdFXOGztScBeKdLVR6r6yefpVxIueROysY6KZfcFWaZPFBfl6p7epBRFCXHvNtMJhMqlYoFCxbg7u4OmJvaPfHEE3z//fc4Oua8ufKzzz5jwoQJOeZv3LgRJyenBwn5gbnbd6AdW1AfXcIWQx0SnUJxVKCCs4ZLqUZGzd7IM5Vt21wuPDzcpscvzqRsrJNyyZ2UTXZpadJ75m3FqW562OyM6bQ9+QEuipHL5ZqzP8oZLq3Jts6xBBW/nFQDKh4NMFE+6Thr1tzpYEg+S9ZJueROysY6KZecCrNuKlAy5O3tjUajyXEVKCYmJsfVotsCAgIIDAy0JEIANWrUQFEULl++TJUqVXJs8+677zJmzBjL66SkJCpUqED79u3x8vIqSMiFwvTnUdRHF9Mm42+M/VaASkVQ3Zs8+dMe9sSqGdunOQ0qlCvyuAwGA+Hh4XTq1AmtVlvkxy/OpGysk3LJnZSNdXFxcbYOodgobnXTw6RZ+QZqfQyKWxB+Ly6gu4N7tuUHom7yzpx9mDDRp34An/epjVpt/lFUPkvWSbnkTsrGOimX3BVm3VSgZEin09GoUSPCw8Pp06ePZX54eDiPP/641W1atWrF4sWLSUlJwcXF3LPM6dOnUavVBAUFWd3G3t4ee/ucTc+0Wq1tToaOH8GJv1Bf3I76wgao1o0mlXx4olEQS/Zf5uM1p1gxopWlIihqNiuXEkDKxjopl9xJ2WQnZXFHsaubHpZjy+Hwb6BSo+r3M1pX72yLT19P5qX5EWQYTLSv5sOXT9ZHq8l5C3KpK5dCIuWSOykb66RccirM8ihwb3Jjxozhl19+YdasWZw4cYLRo0cTFRXFyy+/DJh/ORs0aJBl/YEDB+Ll5cWQIUM4fvw4W7Zs4a233uKFF16w2kSuWCpXAZq/Yp4O/xCMWQC83bUaLvZ2HL6cyOL9l2wYoBBCCFEIEi/DypHm6UfGQHD2nl8T0wwMmb2XxHQDDSqW4/tnGlpNhIQQoqQo8DdY//79mTp1KhMnTqR+/fps2bKFNWvWEBxs7nL62rVr2cYccnFxITw8nJs3b9K4cWOeeeYZevXqxTfffFN476IotB4Djp5w4zQc+BUAX1cHRnU0N/P7ct0pEtPlBjchhBAllMkIy4ZDRiKUbwjt/i/bYkVReGfpYa7cTCfYy4lZzzfBSfdAtx4LIUSx8UDfYiNGjGDEiBFWl82ZMyfHvOrVq5f8m78c3KHdu7D2Ldj0GdR9Cuxdeb5lCL/vvcTZmBT+F36a8Y/VsnWkQgghRMFtnwYXt4HWGfr9AprszVAW7I5i3bFotBoV3w5ogIezzkaBCiFE4ZFr2wXReAh4hkFqrLnSALQaNR/1qgnAvF0XORWdbMsIhRBCiIK7cgA2fmKe7v4leIVlW3wqOplJq8w9xb3TtTp1g8oVcYBCCPFwSDJUEBotdLrVreqO7yDxCgCtq/jQpZYfRpPC+L+OoSiKDYMUQgghCkCfCkuHgikLaj4O9Z/Jtjhdb+S13w6QmWWiXTUfXmgVaqNAhRCi8EkyVFDVe0LFFpCVfudXNOD9HjWxt1Oz83wca45YH4BWCCGEKHbWvQvx58C1PPScCveMGzhx1THOxKTg42rP5Cfr2aznVCGEeBgkGSoolQo6f2yePvgbRB8BoIKnEy+3NTcr+GT1cdL0WbaKUAghhMifEytvdQqkgr4/gpNntsWrD19j4Z5LqFQwtX99vF1ydi0uhBAlmSRDDyKoMdTqCyjw9/twq1ncK+3CCCznyNXEDH7YdM62MQohhBB5SboKf71unm41EkLbZFt8KT6N/1t2GIAR7cJoVdn73j0IIUSJJ8nQg+rwIai1cH4TnN0AgINWw/s9agDww5bzRMWl2TBAIYQQIhcmEyx/GdITIKAetB+XbbHBaOL1hREkZ2TRsGI5RnWsaqNAhRDi4ZJk6EF5hkKz4ebp8A/M4zMAXWv706qyF/osE5NWH7dhgEIIIUQudn4HFzaDnSP0mwl22bvJ/jr8NAcv3cTVwY5pTzeQgVWFEKWWfLv9F63fBIdyEHMcDi4AQKVSMb5XLTRqFeHHr7P5dKxtYxRCCCHudu0wbJhonu76GXhXybZ465lYfthsbur9Rb+6VPB0KuoIhRCiyEgy9F84eUKbt8zT/35i7p4UqOLnyvMtQgCYsPIY+iyTjQIUQggh7qIosGYsmAxQrQc0GpxtcWxyJqMXHUJRYGCzinSvE2CbOIUQoohIMvRfNR0G5YIhJdo89tAtozpVwdtFx/nYVGZvv2DDAIUQQohbTvwFl3aD1gl6TM7WjbbJpPDm4kPcSMmkqp8LH/asacNAhRCiaEgy9F/Z2UPH8ebp7dMg2TzGkJuDlre7VgdgSvhpTlxLslGAQgghBJClh/CPzNMtXgO38tkW/7LtPFtOx+KgVfPdwIY4aDU2CFIIIYqWJEOFoVYfCGwMhlTY+Kll9pONgmhfzQd9lrlXHhl7SAghhM3smwUJF8DZF1q9kW3RwUs3+XLdKQA+7FmLqn6utohQCCGKnCRDheHugVgj5kHMiVuzVUx+sh6+rvacjUlh4krpXU4IIYQNpN+EzZ+bp9u/B/Z3kp3kDANvLIwgy6TQo04AA5pWsE2MQghhA5IMFZbgFlCjFygmCP/QMtvLxZ6p/eujUsHvey+x8tBVGwYphBCiTNr2tXlMIe9q0OA5y2xFUXhv+VGi4tMILOfIp33roLrrPiIhhCjtJBkqTB0ngNoOzvxtHoz1lpaVvXm1XWUA3lt2hEvxMhirEEKIIpJwEXb9YJ7uPAk0dpZFi/ddZuWhq2jUKr4Z0AB3R62NghRCCNuQZKgweYVB4xfN03+/bx7h+5ZRHavQKNiD5MwsXlsYgcEo3W0LIYQoAv9OAmMmhLaBKp0ts8/GJPPRX8cAGNOpKo2CPWwVoRBC2IwkQ4Wt7Ttg7wbRR+DwIstsO42aaU/Xx83BjkOXbjL571M2DFIIIUSZcOUAHFlsnu40ydKVdobByGu/RZBuMNKqshevtA2zYZBCCGE7kgwVNmcvaD3GPP3vJDCkWxYFeTjxRb+6APy42dyFqRBCCPFQKAr8/YF5uu7TUL6+ZdFna05wMjoZL2cd/3uqPmq13CckhCibJBl6GJq9DO4VIOkK7JqebVG3OgE806wiAGP+OERscqYtIhRCCFHanVoLF7eBnQM8+r5l9j/Hr/PrzosATH6qHr5uDraKUAghbE6SoYdB6wiP3vo1buv/ICX7FaAPetakmp8rN1IyGfPHQUwmxQZBCiGEKLWMhjs9mzZ/BcqZu8tO1xst9wkNfSSU9tV8bRWhEEIUC5IMPSx1noSAeqBPho2fZFvkoNXw7cAGOGjVbD1zg5+2nrdRkEIIIUqlA79C3Blw8oJHRltm/7D5HFduplPe3YE3O1ezYYBCCFE8SDL0sKjV0PlWErR/NpwJz7a4qp8rH/WqBcDk9aeIiEoo6giFEEKURhlJsOnWAKvt3gUHdwAuxafxw+ZzAIzrURNHncZWEQohRLEhydDDFNoamr5knl4xIkdzuaebVKBHnQCyTApv/B5BUobBBkEKIYQoVbZPg9RY8KoMjQZbZn+8+jiZWSZaVPKiex1/28UnhBDFiCRDD1unieBTA1Jj4K/XzL373KJSqfi0bx0CyzlyKT6d95YdQVHk/iEhhBAPKPEK7PzOPN1xAmjMg6huPRPL+mPX0ahVTHi8FiqV9B4nhBAgydDDp3WEfr+ARgen18G+mdkWuztq+WZAAzRqFasOX+OPfZdsFKgQQogSb+MnkJUBFVtA9R4AGIwmJqw8DsCgFsFU9XO1ZYRCCFGsSDJUFPxrm3+hA1g/DmKzD7jaKNiDNztXBeCjv45xNia5qCMUQghR0l07DAd/M093/tgywOqvOyI5G5OCl7OOUR2r2jBAIYQofiQZKirNXoawR82/2C15EbKyjy/0cpswHqnsTYbBxGu/RZBhMNooUCGEECWOokD4B4ACtftBUGMAYpIzmPrPGQDe7loNd0etDYMUQojiR5KhoqJWQ+8Z5m5Orx+BDRPvWazi66fq4eWs42R0Mp+sPmGjQIUQQpQ4ZzfA+U3mJtkdPrTM/nLdKVIys6gb5M6TjSrYLj4hhCimJBkqSq7+8NitG1t3fgfnNmZb7OvmwJSn6gEwb9dF1h2NLuoIhRBClDQm462rQph7MPUIAeBAVAJL9l8GYMJjtVCrpdMEIYS4lyRDRa16d2j8gnl6+cuQFp9tcbtqvrzUphIA7yw9zJWb6UUdoRBCiJLk4AKIOQ4O5aDNWABMJoXxfx0D4IlGQTSo6GHDAIUQoviSZMgWOn8CXlUgJRr+ej1bd9sAYztXo16QO4npBkYujCDLaLJRoEIIIYq1zBT499YA323fBkdz0rN4/yUOX07E1d6Od7pWt2GAQghRvD1QMjR9+nRCQ0NxcHCgUaNGbN26NV/bbd++HTs7O+rXr/8ghy09dE7wxExQa+HkKjgwN/tiOzXfDGiAi70d+y4m8NFfx2T8ISGEEDnt/M78w5pHCDQZCkBiuoEv15l7LR3ZsQo+rvY2DFAIIYq3AidDixYtYtSoUYwbN46IiAhat25Nt27diIqKynO7xMREBg0aRIcOHR442FIloN6dm1zX/R/cOJNtcbCXM189UReVChbsjuKbDWdtEKQQQohiKzkatn9jnu44HuzMSc//wk8Tl6qnsq8Lz7cMsVl4QghREhQ4Gfr666958cUXGTp0KDVq1GDq1KlUqFCBGTNm5Lnd8OHDGThwIC1atHjgYEudFq9BaBswpMHSoZClz7a4W50AxveqBcD//jnN/F0XbRGlEEKI4mjjp2BIhaAmULM3AKeik5l3q64Y36sWWo20hhdCiLwU6FtSr9ezf/9+OnfunG1+586d2bFjR67bzZ49m3PnzvHRRx89WJSllVoNfX403/R67SBs+jTHKs+3DOH1RysD8MGfR1l75FrRxiiEEKL4iTkBEfPM07cGWFUUc6cJRpNC11r+PFLF27YxCiFECWBXkJVv3LiB0WjEz88v23w/Pz+io613A33mzBn+7//+j61bt2Jnl7/DZWZmkpl5Z1DSpKQkAAwGAwaDoSAhF3+OPqh6TMVu6WCUbVMxhrRFCX4k2yqvtwslJimDRfsu88bvEcy2V9Ms1NNSFqWuTAqBlI11Ui65k7KxTsrjjuJUN2nWv49aMWGq3gtjQCMwGFh7NJqd5+Owt1PzTpcqNvvbyWfJOimX3EnZWCflkrvCLJMCJUO3qVTZxypQFCXHPACj0cjAgQOZMGECVatWzff+P/vsMyZMmJBj/saNG3Fycip4wMWemnpebQmJ24xh0QtsrP4JBjvnbGs0t4MTnmoOx6sZ+uteXq9lJOjWKuHh4TaIuWSQsrFOyiV3UjbZpaWl2TqEYqO41E0eqWdoc+4fTGj4V9Wa1DVr0Bvhk4MaQEV7/ywO79zI4SKLyDr5LFkn5ZI7KRvrpFxyKsy6SaUUoJsyvV6Pk5MTixcvpk+fPpb5I0eO5ODBg2zevDnb+jdv3sTDwwONRmOZZzKZUBQFjUbD33//zaOPPprjONZ+fatQoQLXrl3Dy8urQG+wxNCnYDfzUVTx5zHVeBxjn1/gngQz02BkyNwD7I1MwNtFx4IhDTm5bxudOnVCq9XaKPDiyWAwEB4eLmVzDymX3EnZWBcXF0dAQACJiYm4ubnZOhybKi51k2b5MNTHl2OqOxBjL3MHClM3nOX7TecJLOfAujda4aDV3GcvD498lqyTcsmdlI11Ui65K8y6qUBXhnQ6HY0aNSI8PDxbMhQeHs7jjz+eY303NzeOHDmSbd706dP5999/WbJkCaGhoVaPY29vj719zq5AtVpt6T0ZtB7Q7xeY2Rn1iT9RV+0CDZ7JvopWyy/PN6H/jzs5GZ3MsAWHGBZaysvlP5KysU7KJXdSNtlJWdxRLOqmpKtwciUA6havoNZqiYpL4+dtkQB80LMmrk4ORRPLfchnyTopl9xJ2Vgn5ZJTYZZHgbuZGTNmDL/88guzZs3ixIkTjB49mqioKF5++WUA3n33XQYNGmTeuVpN7dq1sz18fX1xcHCgdu3aODs753WosiewEbR/zzy99m2IO5djFXdHLXNfaEoFT0ei4tP58aSG5IysIg5UCCGETeybBaYsCG4FAXUB+Hj1cfRZJlpV9qJLLX8bByiEECVLgZOh/v37M3XqVCZOnEj9+vXZsmULa9asITg4GIBr167dd8whkYdWoyD4EdCnwLKXwJjzBjFfNwfmvtAMT2ctl1NVvLrwIJlZxqKPVQghRNExZMC+2ebpZsMB2HI6lr+PX8dOrWJ8r1pW798VQgiRuwcagGDEiBFERkaSmZnJ/v37adOmjWXZnDlz2LRpU67bjh8/noMHDz7IYcsGtQb6/ggO7nBlH2z+0upqod7OzHyuEfZqhZ3n4xmz6BBGU75v/xJCCFHSHF0KaTfAvQJU64E+y8T4lccA8zAMVfxcbRygEEKUPDIaW3HkHgQ9p5qnt06GcxutrlY70I0Xq5nQalSsPnKN8X8dowD9YQghhCgpFAV2/2CebjIUNHb8uiOS87GpeLvoGNmxim3jE0KIEkqSoeKqdl+o/ywoJlj0LFyNsLpatXIKk/vVQaWCebsu8u2/Z4s4UCGEEA9d1C6IPgx2jtBwEDFJGUzbcAaAt7tWx81Bbq4WQogHIclQcdbzawhtY75/aP4TcMN6otO9jj8THqsFwNfhp1mw+2JRRimEEOJhu31VqO5T4OTJ5+tOkpKZRf0K5XiiYZBtYxNCiBJMkqHizM4env4NAuqb24nP62PuVtWKQS1CeOPRygB8sOIo645eK8JAhRBCPDSJl+GEuTttmg3n6JVElh24AsCEx2qhVkunCUII8aAkGSru7F3hmSXgVRkSo2BeX0iLt7rq6E5VGdC0IiYF3lh4kJ3n4oo4WCGEEIVu70xQjBDSGvxqMX2TuZXA4/XLU69COdvGJoQQJZwkQyWBiw88uwxcAyD2BPzWH/RpOVZTqVR83Ls2XWr5oTeaeGnuPo5dTbRBwEIIIQqFIR32zzFPN3uZc7EprD0aDcCIdpVtF5cQQpQSkgyVFB7B5oTIoRxc3gOLn7c6BpFGrWLa0w1oFupJcmYWz8/aw5HLkhAJIUSJdGQJpMeDe0Wo1o0fN59DUaBjDV+q+UtX2kII8V9JMlSS+NWEgX+YexM68zeaVW+Ye5u7h4NWw8/PN6Z2oBs3UvT0/2knG0/F2CBgIYQQD+zu7rSbDuNqkt5yr9CI9nJVSAghCoMkQyVNxWbw1FxQ26E+upjaVxaaK8x7uDloWTisOa2reJOmNzL01338sfeSDQIWQgjxQC5uh+tHQesEDZ/j563nyTIpNK/kScOKHraOTgghSgVJhkqiqp3h8ekAhMWuR71jmtXVXB20zBrchL4NAzGaFN5eepip/5yWgVmFEKIksHSn3Z94kzO/7zH/oCX3CgkhROGRZKikqtcfY6ePAdBs+hj2/2p1Na1GzZQn6/HarSYVU/85w/8tPYLBmLN5nRBCiGLiZhScXG2ebjacOdsvkG4wUjvQjdZVvG0bmxBClCKSDJVgpqYvc9qvl/nFqlFw/C+r66lUKsZ2qcYnfWqjVsGifZcYNncfqZlZRResEEKI/Nv7i/me0ErtSHarzJwdkQC82q4yKpWMKySEEIVFkqES7kTAE5jqP2uuNJe+CBe25LruM82C+em5xjho1Ww6FcvTP+0iNjmzCKMVQghxX/q0O1f7m73Mb7ujSMrIopKPM11q+ds2NiGEKGUkGSrpVCqM3SZD9Z5g1MPCgXD1YK6rd6zpx8JhzfF01nHkSiJ9Z2znfGxK0cUrhBAib0f+gIyb4BFCRkgHftl2AYCX24ahVstVISGEKEySDJUGajvoN9M8Ork+Geb3g7hzua7eoKIHS19pSbCXE5fi0+k3Ywf7LyYUYcBCCCGsUhTY/aN5uulLLD14jdjkTALcHehdP9C2sQkhRCkkyVBpoXWAp38D/7qQdgPm9Yaka7muHurtzNJXWlIvyJ2ENAMDf97F+mPRRRevEEKInCK3Qsxx0DqTVXcgP2w2/7D1UptK6OykyhZCiMIm36yliYMbPLsUPCuZeyKa3w/Sc7/i4+1iz8KXmtOhui+ZWSZemb+feTsjiy5eIYQQ2e261Z12/QGsPpPGpfh0PJ11PN2kom3jEkKIUkqSodLGxReeWw4u/hBzDBY8BWnxua7upLPjx+caMaBpRUwKfPDnMT5fexKTScYiEkKIIpUQCafWAKA0GcaMTearQkNahuCo09gwMCGEKL0kGSqNPELguWXg4A6X98AvHfO8h8hOo+bTPrV5s1NVAH7YfI4xfxxEnyVjEQkhRJHZ8zOgQNij/BvnwcnoZJx1Gga1CLF1ZEIIUWpJMlRa+dWCIWvBvQLEn4NfOkDktlxXV6lUvN6hCl89URc7tYoVB68yePYekjIMRRi0EEKUUZkpcGAeAErT4Xy/8SwAz7YIxt1Ja8vIhBCiVJNkqDTzqwVDN0BgY/O9Q3N7Q8SCPDd5snEFZg5ugrNOw45zcfT8ZhsHoqSnOSGEeKgOL4LMRPAIZbddIw5E3URnp+bFR0JtHZkQQpRqkgyVdq5+MHgV1OoDJgP8OQL+GQ+m3JvAta3qw6LhLQgs50hUfBpP/rCTbzacwSj3EQkhROG7uzvtZsOZvtk8rtCTjYLwdXWwYWBCCFH6STJUFmgdod8saPOW+fW2/8Hi582jnOeidqA7a0a2ple98hhNCl+Hn+bpn3ZyKT73bYQQQjyA85vgxinQuXDcrxdbTseiVsHwNmG2jkwIIUo9SYbKCrUaHn0f+vwIGh2c+Avm9IDk3McWcnfU8s3T9fn6qXq42NuxNzKB7tO28ufBK0UYuBBClHK3rwrVf4bvdlwH4LF65ano5WTDoIQQomyQZKisqfc0DPoTHD3h6gH4uQNEH8l1dZVKRd+GQax5ozUNK5YjOTOLkb8fZPSig9K5ghBC/Ffx5+H0OgCiqjzD2qPmH6heaVfZllEJIUSZIclQWRTcEob+A15VIOkyzOoKp9fnuUlFLyf+GN6CkR2qoFbB8ogrdJ+2lX2RuY9hJIQQ4j5ud6dduRPfHTLfPtSxhi/V/F1tHZkQQpQJkgyVVV5hMDQcQtuAPgUWPm0e+VzJvZMEO42a0Z2qsvjlFgR5OHI5IZ2nftzJ/8JPk2WUMYmEEKJAMpMhYj4AcbWHsDzC3ARZrgoJIUTRkWSoLHP0gGeXQcNBoJhg3TuwZiwYs/LcrFGwJ2tGtqZvg0BMCkzbcIanftxJVJx0riCEEPl26HfITAKvynwXVRGDUaFZqCeNgj1sHZkQQpQZkgyVdRot9PoGOk0CVLD3F/jtKchIzHMzNwctX/evz7Sn6+PqYMeBqJt0/2Yryw5cRsnj6pIQQgjMwxvc6jghtf6L/L7XfFXo1fZyVUgIIYqSJEMCVCpo9Qb0nw9aJzi3AWZ2hoSL99308fqBrB3ZmiYhHqRkZjHmj0O88ftBEtOlcwUhhMjV+X8h7gzoXJmV0px0g5HagW60ruJt68iEEKJMkWRI3FGjJwxZC64BEHsSfn4UIrfdd7MgDyd+f6kFYztXRaNWsfLQVbpP28qeC9K5ghBCWLX7JwD0dQfy8+5YAEa0q4xKpbJlVEIIUeY8UDI0ffp0QkNDcXBwoFGjRmzdujXXdZctW0anTp3w8fHBzc2NFi1asH593j2XCRsqXx+G/Qv+dSHthnksojVvQWZKnptp1Cpee7QKS15uQbCXE1duptP/p53839LDxCZnFk3sQghREqTFw9l/AFim6UJSRhaVvJ3pUsvfxoEJIUTZU+BkaNGiRYwaNYpx48YRERFB69at6datG1FRUVbX37JlC506dWLNmjXs37+f9u3b06tXLyIiIv5z8OIhcStvvkLUcJD59Z6fYEZLOL/5vps2qOjB6jda81TjIBQFft97ifaTN/HD5nNkZhkfcuBCCFECnFgJihGTXx2mHDDfY/lyuzA0arkqJIQQRa3AydDXX3/Niy++yNChQ6lRowZTp06lQoUKzJgxw+r6U6dO5e2336ZJkyZUqVKFTz/9lCpVqrBy5cr/HLx4iOxd4LFv4bnl4F4Bbl6EuY/BylGQkZTnpi72dnz5RD2WvtKCukHupGRm8fnak3T+3xbWH4uWDhaEEGXb8RUAHHZvT2xyJgHuDvSuH2jbmIQQooyyK8jKer2e/fv383//93/Z5nfu3JkdO3bkax8mk4nk5GQ8PT0LcmhhK2GPwoidEP4R7JsJ+2fDmXB4bBpU7pjnpo2CPVkxohXLI67wxbqTXIxLY/i8/bSo5MWHvWpSI8CtiN6EEEIUE6lxlqvsX0TVAGBY60ro7OQWXiEKymg0YjCU3g6bDAYDdnZ2ZGRkYDSWrdY1Wq0WjUZTJMcqUDJ048YNjEYjfn5+2eb7+fkRHR2dr31MmTKF1NRUnnrqqVzXyczMJDPzzn0mSUnmKxEGg6FUn/QFdbssHnqZqB2gyxeoqvVEs3oUqpsXYX4/TPWewdhxIji457n5Y3X96FDNix+3XmDm9ovsPB9Hj2+28lTjIEZ1qIyXs67QQy6ysilhpFxyJ2VjnZTHHYVRN6mOrcBOMZLgVoOdMe54OGnp18C/VJWzfJask3LJXUHLRlEUYmJiLJ/B0kpRFPz9/YmKiiqTnau4ubnh6+tr9b0X5ueoQMnQbfcGpShKvv5ICxcuZPz48fz555/4+vrmut5nn33GhAkTcszfuHEjTk5OBQ+4lAsPDy+yY2kqvk8N7WIqxYajPrSAzONrOFRhMNfdG9x32+rA/9WBlVFqIuLU/L73MisOXKJLkIk2/goP44fRoiybkkTKJXdSNtmlpclgyrcVRt3U4uwsfIFFyXXMr70y2PTP34UZZrEhnyXrpFxyl9+ycXV1xcPDA29vb3Q6XZlMFEozRVHQ6/XExsZy+vRpkpOTc6xTmHWTSinADRx6vR4nJycWL15Mnz59LPNHjhzJwYMH2bw59xvsFy1axJAhQ1i8eDE9evTI8zjWfn2rUKEC165dw8vLK7/hlnoGg4Hw8HA6deqEVqst0mOrLu1Cs+oNVPHnATDVfhJj50/BMX8jp++NTOCTtSc5dtV8god4OfF/XavyaDWfQvlSs2XZFGdSLrmTsrEuLi6OgIAAEhMTcXMr201b/3PdlBaH3dSaqBQjbTO/5qo6gO1vt8XDqfCvjtuSfJask3LJXUHKxmg0cv78eXx8fEr9/4SKopCcnIyrq2uZTPji4uKIjY2lUqVKOZrMFWbdVKArQzqdjkaNGhEeHp4tGQoPD+fxxx/PdbuFCxfywgsvsHDhwvsmQgD29vbY29vnmK/VauULxAqblEul1vDydtj4CeyajvroYtQXNkPPr6FGr/tu3rKKL3+F+bB0/2W+XH+KyLg0Xl5wkNZVvHm/R02q+bsWSphyzlgn5ZI7KZvspCzu+M9109l1oBi55liVixn+9Kjlj6+780OItHiQz5J1Ui65y0/ZGI1GVCoVLi4uqNWl+147k8kEmFtklfb3ao2Liws3btwActZFhfkZKnDJjhkzhl9++YVZs2Zx4sQJRo8eTVRUFC+//DIA7777LoMGDbKsv3DhQgYNGsSUKVNo3rw50dHRREdHk5iYWGhvQtiIzgm6fAIv/A3eVSE1BhY9C4uHQOqN+26uUat4qkkFNr3VjlfahaHTqNl65gbdpm3hgxVHiU/VF8GbEEKIInJsOQB/ZDQG4KnGFWwZjRAlWlm8UlLWFNXfuMDJUP/+/Zk6dSoTJ06kfv36bNmyhTVr1hAcHAzAtWvXso059OOPP5KVlcWrr75KQECA5TFy5MjCexfCtio0geFb4ZHRoNLAsWXwfTM4ugzy0QrTxd6Od7pW558xbelayx+TAvN2XeSRL/5l4srjXLmZXgRvQgghHqLUG3BhCwBLM5oQ4O7AI5W9bRyUEEKIB7rmNmLECCIjI8nMzGT//v20adPGsmzOnDls2rTJ8nrTpk0oipLjMWfOnP8auyhOtA7QcTwM/Qd8a0LaDVgyBGZ2gnMb85UUVfRy4ofnGrFwWHNqB7qRpjcya/sF2n65kTGLDnIyunT3GiOEKMVOrATFRKS2ClGKH/0aBskgq0KIApkzZw7lypWzdRilTtlrgCgersCG8NJmaPt/YOcIl/fCvN4wpydczN9YVC3CvFj52iPMfaEprSp7kWVSWBZxha5TtzJ49h52nouTgVuFECXLrSZyv6c3AuCJRkG2jEYIYSOXLl3ixRdfpHz58uh0OoKDgxk5ciRxcXHZ1gsJCWHatGk2irJwtGvXjlGjRtk6jPuSZEgUPjsdtH8XRh6CZi+DRgcXt8HsbjC3N1zed99dqFQq2lT1YcHQ5qx87RF61g1ArYJNp2IZ8PMuen+/nTVHrmE0SVIkhCjmUm9A5FYAVhub0TTUkxDv0ttxghDCuvPnz9O4cWNOnz7NwoULOXv2LD/88AMbNmygRYsWxMfH2ySusj72lSRD4uFx9YNuX8AbEdBoCKjt4PxG+KUD/NYfrh3K127qBLnz3cCGbBrbnueaB2Nvp+bQ5URGLDjAo1M2MX/XRTIMZWtkZiFECXLiL1BMnFKHcUnx40m5KiREmfTqq6+i0+n4+++/adu2LRUrVqRbt278888/XLlyhXHjxgHmKyoXL15kzJgxeHh45OhWev369dSoUQMXFxe6du3KtWvXsi2fPXs2NWrUwMHBgerVqzN9+nTLssjISFQqFX/88Qft2rXDwcGB+fPnW4335s2bvPTSS/j5+eHg4EDt2rVZtWoVYO7aesCAAQQFBeHk5ESdOnVYuHChZdvBgwezefNmpk2bhkqlQqVSERkZSUJCAs888ww+Pj44OjpSpUoVZs+eXSjl+6AeaNBVIQrEPQh6TYVHRsHmr+DQb3B6nflRoxe0ew/8at53NxW9nJjUuzajOlbh150Xmbszkotxaby/4ij/Cz/N4JYhPNcimHKlbMwOIUQJd2wFAMsym+Ks09C9ToBt4xGilFEUhXQb/SjqqNXkq9ez+Ph41q9fzyeffIKjo2O2Zf7+/jzzzDMsWrSI6dOns2zZMurVq8ewYcPo378/rq53hhtJS0tj8uTJzJs3D7VazbPPPsvYsWNZsGABAD///DMfffQR3333HQ0aNCAiIoJhw4bh7OzM888/b9nPO++8w5QpU5g9e7bVIQNMJhPdunUjOTmZ+fPnExYWxvHjxy2JWUZGBo0aNeKdd97Bzc2N1atX89xzz1GpUiWaNWvGtGnTOH36NLVr12bixIkA+Pj4MHLkSI4fP87atWvx9vbm7NmzpKfbtqMsSYZE0fEIgd7fm3ud2/w5HFlivqn4xCqo3Q/avQvele+7Gy8Xe8Z0qsrLbSvxx95L/Lz1AldupjMl/DQzNp+jf5MKDG4uXdYKIYqBlNg7TeRMzejZoDzO9lL1ClGY0g1Gan643ibHPj6xC066+3+mz5w5g6Io1KhRw+ryGjVqkJCQQGxsLL6+vmg0GlxdXfHz88s2qKjBYOCHH34gLCwMgNdee82SbABMmjSJKVOm0LdvXwBCQ0M5fvw4P/74Y7ZkaNSoUZZ1rPnnn3/Ys2cPJ06coGrVqgBUqlTJsjwwMJCxY8daXr/++uusW7eOxYsX06xZM9zd3dHpdDg5OeHv729ZLyoqigYNGtC4sXmIgZCQkPuW3cMm38ii6HlXhn6/QOs3YdNncPxPOLrE3CV3vQHQ9m1z4nQfTjo7BrcK5dnmwaw+co0fN5/n+LUkZm+PZO7Oi9RwV6MJvk7n2uXR2UmLUCGEDdxqIndEqcRlxZcnG0sTOSFETrc7hrrfVSYnJydLIgQQEBBATEwMALGxsZYOGoYNG2ZZJysrC3d392z7uZ2M5ObgwYMEBQVZEqF7GY1GPv/8cxYtWsSVK1fIzMwkMzMTZ+e874d85ZVX6NevHwcOHKBz58707t2bli1b5rnNwybJkLAd3xrw1Fy4dhg2fgqn18LBBXB4ETR41pwslat4393YadQ8Xj+Qx+qVZ9vZG/y4+Tzbzt7gaIKa134/RDmn4zxWrzz9GgZRN8hdBmoTQhSd4ysAWJXVjErezjQK9rBtPEKUQo5aDccndrHZsfOjcuXKqFQqjh8/Tu/evXMsP3nyJB4eHnh75z3+mFarzfZapVJZEimTyQSYm8o1a9Ys23r33nd0v6Tl3qZ895oyZQr/+9//mDp1KnXq1MHZ2ZlRo0ah1+vz3K5bt25cvHiR1atX888//9ChQwdeffVVJk+enOd2D5MkQ8L2AurCwN/h8n7Y+Amc2wD758CBuVC5EzQaDFU6gybv01WlUtG6ig+tq/hw/EoCU5Zu42iyI9eTM5m78yJzd16kiq8LfRsG0adBIP7uDkXy9oQQZVRKLERuA8xN5AY2DpIfY4R4CFQqVb6aqtmSl5cXnTp1Yvr06YwePTpbshEdHc2CBQsYNGiQ5TtCp9NhNBbsPig/Pz8CAwM5f/48zzzzzH+Kt27duly+fJnTp09bvTq0detWHn/8cZ599lnAnIidOXMmWzPA3N6Dj48PgwcPZvDgwbRu3Zq33nrLpsmQtB0SxUdQI3huGQxZB6FtQTHBmfXw+wCYWhv+/QRuRuVrV1V8XXgs2MTmsW349YWmPFavPPZ2as7EpPDFupO0/HwDz83czZ8Hr5Cul57ohBAPwa0mcgdNlbiKL/0aShM5Icqy7777jszMTLp06cKWLVu4dOkS69ato1OnTgQGBvLJJ59Y1g0JCWHLli1cvXqVGzdu5PsY48eP57PPPrN0YHDkyBFmz57N119/XaBY27ZtS5s2bejXrx/h4eFcuHCBtWvXsm7dOsB8pSs8PJwdO3Zw4sQJhg8fTnR0dLZ9hISEsHv3biIjI7lx4wYmk4kPP/yQP//8k7Nnz3Ls2DFWrVqV631URUWSIVH8BLeA5/+C1/ZDy9fByQuSr8GWL2FqXZj/BJxcDcas++5Ko1bRtqoP3wxowN73O/J53zo0CfHApMDWMzcY+ftBmnzyD+8sOcyeC/EymKsQovDcGmh1jbEZ7ar54ucmV6OFKMuqVKnCvn37CAsLo3///oSFhfHSSy/Rvn17du7ciaenp2XdiRMncvHiRRo2bIifn1++jzF06FB++eUX5syZQ506dWjbti1z5swhNDS0wPEuXbqUJk2aMGDAAGrWrMnbb79tudLzwQcf0LBhQ7p06UK7du3w9/fP0fxv7NixaDQaatasiY+PD1FRUeh0Ot59913q1q1LmzZt0Gg0/P777wWOrTCplBLw319SUhLu7u7cuHEDLy8vW4dTbBgMBtasWUP37t1ztCEtVbIy4eQqc9O5C1vuzHfxh4bPQYPnwCM42yb3K5uLcaksPXCFZQcucznhTpeOFT2d6NswkL4Ngqjo5fSw3pHNlJlz5gFI2VgXFxeHt7c3iYmJ2Xo0Evepm1JiUKZUQ6WYeCRzGuMGdqFbGelSWz5L1km55K4gZZORkcGFCxcIDQ3FwaF0/8BgMplISkrCzc0NtbrsXb/I629dmHVT8W5gKQSAnb256+3a/SDuHBz4FSIWQEo0bPkKtkyGyh3M9xZV7Qqa+1cywV7OjOlUlVEdqrAnMp6l+y+z5sg1ouLTmPrPGab+c4Zqfq48WsOXjjV8qV/BA41a2voLIfLpxF+objWRS3MKpEON/P+yK4QQouhIMiRKFq8w6DQR2r8Pp1abrxad3wRn/zE/XPzMPdHVHZiv3anVKppX8qJ5JS8mPF6L9ceiWbr/CjvPx3HqejKnriczY9M5PJ11tKvmQ4fqfrSp6o2rg/yqJ4TIw62BVlcbm/N4E+neXwghiitJhkTJZKeDWn3Mj/jz5p7nIuZDynXYOgXt1im0dgpDXe4M1HwMfKz3k383J50dfRoE0adBEDfT9Gw+HcuGEzFsOhVDfKqeZQeusOzAFbQaFU1DPelQ3Y8ONXwJ9sq7e0ohRBmTfB3l4nZUmO8X+rmRDAIthBDFlSRDouTzrAQdx0O798xjFe2fA+f+xTPtHGycZH54VYZq3aF6TwhqDOq8xwUo56Tj8fqBPF4/EIPRxP6LCWw4cZ0NJ2M4H5vK9rNxbD8bx8RVx6ns60KH6r50qOFHw4rlsNPIL8BClGmWJnJheAZWpmZ5uddKCCGKK0mGROlhp4Oaj0PNxzHER3F8+RTqaC+hjtwCcWdhxzfmh7MPVOsG1XpApbagzXtgMa1GbWlKN65HTS7cSGXDiev8ezKGPRfiORuTwtmYFH7ccp5yTlraVTWPddQ01JMgD0cZV0SIMkY5vgIVsMrYnCcbS3faQghRnEkyJEon1wAivR+lZvfuqI3p5vuJTq2B039Daqy5Wd2BuaB1Mne+UK0HVO0CTp733XWotzNDW1diaOtKJKYb2HrG3Jxu46kYbqYZWHHwKisOXgUgwN2BJiGeNA01Pyr7uKCWjhiEKL2Sr0PkdgD+UTVnRb3yNg5ICCFEXiQZEqWfgxvU7mt+ZOnh4nZzYnRyNSRdgRMrzQ+VBoJbmpvTVesKHqFwn6s67o5aetYtT8+65TGaFA5EJfDvyRh2n4/j8OVEriVm8Nehq/x1yJwceThpsyVHNQPcpFmdEKXJib9QoRBhqkztmrUp56SzdURCCCHyIMmQKFvsdBDW3vzo9iVcO3QnMbp+FCK3mh/r3wXX8ubk6PbDuxrk0c+/Rq2iSYgnTULMV5fS9UYiohLYExnPngvxHIhKICHNwN/Hr/P38esAOOs0NAz2oOmtBKlehXI4aPO+n0kIUXyZji5HDaw2NuOpxtJxghBCFHeSDImyS6WC8vXNj/bvQUIknFprToyidkHyVTi6xPwAcPQ0J0UVW5if/euCJvePkKNOQ8vK3rSs7A2APsvE0auJ7L1gTo72RsaTlJHF1jM32HrmBgA6jZq6Qe7UCXKndnl3agW6UdnHRa4eCVESJEejitoBwAHn1rx767MvhBCi+JJkSIjbPEKg+Svmhz4NruyDizvMj0t7ID0eTq4yPwB0LlCh6a0rR62gfEPQ5j4ats5OTcOKHjSs6MHwtmGYTAqnriezNzKe3bcSpNjkTPZdTGDfxQTLdvZ2aqr7u1Ir0J1a5d2oXd6dav6ucgVJiOLmxEpUKBwwVaZV44YyULMQosAiIyMJDQ0lIiKCunXrFso+VSoVy5cvp3fv3oWyv9JGkiEhrNE5QWgb8wPM9xpdO2S+3yhqJ1zcCZmJcO5f8wNAo4PAxhDcAoKagH8dcAvM9b4jtVpFjQA3agS4MahFCIqicDEujf0XEzh6NZFjV5M4fjWJlMwsDl1O5NDlRMu2GrWKKr4u1Cp/K0EKdKdGgKsMBiuEDWUeWoI95iZygxpJL3JCiOwGDx7Mr7/+annt6elJkyZN+PLLLwst8SkONm3aRPv27UlISKBcuXK2Due+JBkSIj/sdFChifnBKDAZIea4OSm6uN189Sg1BqJ2mB+3OXqYkyL/uuBX2zztUw00OZMWlUpFiLczId7O9Lv1j5TJpBAVn2ZJjo5eMT/Hp+o5GZ3Myehklh64s49Qb2dqBLgS5uNCmI8LlXycqeTjgou9fNSFeKiSo9Fd2Q3AtcAuMhizEMKqrl27Mnv2bACio6N5//336dmzJ1FRUTaOrOyS/5CEeBBqza0kpw40ewkUBeLP30qMdpqvIsWehPQEuLDF/LhNowOf6uYE6fY+/GuDg3vOw6jvJEg965q76FUUheikDI5dSbIkSceuJHI1MYMLN1K5cCM1x3783RwI83Um1MuJ9Osq3M7GUS3AnQB3BxkHSYhCoBz/09JErkOzRrYORwhRTNnb2+Pv7w+Av78/77zzDm3atCE2NhYfH58c6xuNRoYOHcrGjRuJjo6mYsWKjBgxgpEjR2Zbb9asWUyZMoWzZ8/i6elJv379+O6776zGMHHiRL7//nvWr19P/fr1ra7z119/MXHiRI4ePYqLiwtt2rRh2bJlAMyfP5+pU6dy6tQpnJ2defTRR5k6dSq+vr5ERkbSvn17ADw8PAB4/vnnmTNnDkuWLGHChAmcPXsWJycnGjRowJ9//omzs21/PJJkSIjCoFKBV5j50XCQeZ4hw5wQRR8x91QXfcT8yEyC6MPmx93KBd9Jjrwqm/flGWbuGjzboVQEuDsS4O5Ix5p+lvnxqXqOXU3k9PUUzsWmcC4mhXOxqdxIySQ6KYPopAy2n40DNCyN3A+Ak05jvnrkbb6SFObrTLCnM0EejpRz0kqiJEQ+Je9fjBsQrmrJ63X8bR2OEGWLooAhzTbH1jrddxiO3KSkpLBgwQIqV66Ml5eX1XVMJhNBQUH88ccfeHt7s2PHDl566SUCAgJ46qmnAJgxYwZjxozh888/p1u3biQmJrJ9+/Yc+1IUhVGjRrFixQq2bdtGlSpVrB5z9erV9O3bl3HjxjFv3jz0ej2rV6+2LNfr9UyaNIlq1aoRExPD6NGjGTx4MGvWrKFChQosXbqUfv36cerUKdzc3HB0dOTatWsMGDCAL7/8kj59+pCcnMzWrVtRFOWByq4wSTIkxMOidbjTW91tigI3L95JjG4/Ei+Z59+8eKeDhtucfe8kRl6VzImSZxh4VjLf23SLp7OO1lV8aF0l+y9LiekGzseaE6Mz0UnsOHqOVI0rUfFppOmNHL2SxNErSTnCd9ZpCPJwIsjD8dbDKduzJEtC3JIcjUvMPgBM1R/DSSdVqxBFypAGn9pogOP3roIu/1c2Vq1ahYuLCwCpqakEBASwatUq1LkM3aHVahk/frxleWhoKDt27OCPP/6wJEMff/wxb775ZrarRU2aNMm2n6ysLAYNGsS+ffvYvn07QUG539f4ySef8PTTTzNhwgTLvHr16lmmX3jhBct0pUqV+Oabb2jatCkpKSm4uLjg6WkeYsTX19dyz9C5c+fIysqib9++BAcHA1CnTp28C6uIyDe2EEVJpTL3WucRAjV63ZmfFg/Xj926inQM4s9B3FlIjTXfi5QaY+644V6u5e9ckfK89ewRCu6BlmZ37o5aGlT0oEFFDwwGA2uyztC9eytQa7gUn8a52NS7riSlcDkhnZjkTFL1Rk5dT+bU9WSrb8VashTo4YifmwN+bvb4uNpjbyc93onSz3BiNWoU9puq0LllY1uHI4Qoxtq3b8+MGTMAiI+PZ/r06XTr1o09e/ZYkoR7/fDDD8yaNYuLFy+Snp6OXq+3NG+LiYnh6tWrdOjQIc/jjh49Gnt7e3bt2oW3d97d/h88eJBhw4blujwiIoLx48dz8OBB4uPjMZlMAERFRVGzZk2r29SrV48OHTpQp04dunTpQufOnXniiScsTelsSZIhIYoDJ08IbW1+3C0j6VZidOthmT4LGTfNYyElXzUPFHsvnas5KXILvPUchMrFH5+kyxBXBa1nCJV8XKjk40In/LIf1mDk6s10LieYH5cS0m5Nm59j85Esgflqla+rPb5uDvi52t+VKJmf/dwc8HG1RyvjKIkSLPXwSgB2ObZhRMVytg1GiLJI62S+QmOrYxeAs7MzlStXtrxu1KgR7u7u/Pzzz3z88cc51l++fDlvvvkmU6ZMoUWLFri6uvLVV1+xe7e5wxZHR8d8HbdTp04sXLiQ9evX88wzz+S5bl77TE1NpXPnznTu3Jn58+fj4+NDVFQUXbp0Qa/X57qdRqMhPDycHTt28Pfff/Ptt98ybtw4du/eTWhoaL7ew8MiyZAQxZmDG5RvYH7cKy3+rgTp7J3phIvmREmfbL5nKfakZRM7oCXAuS/NMxw9wC0oe9Lk4o+Dsw+VXHyo5OcDlfzAzj7boTMMRq5YkqU0S9J0JSGNmORMYpIy0RtNxKfqLT3f5UalAi9nHT6u5sTIy1mHh5MOLxcdns7ZH17OOtwctKhl/BZRjHjdPAz24NbwCWk6KoQtqFQFaqpWnKhUKtRqNenp6VaX79y5k5YtWzJixAjLvHPnzlmmXV1dCQkJYcOGDZaOC6x57LHH6NWrFwMHDkSj0fD000/num7dunXZsGEDQ4YMybHs5MmT3Lhxg88//5wKFSoAsG/fvmzr6HQ6wNz5w73vtVWrVrRq1YoPP/yQ4OBgli9fzpgxY3KNpShIMiRESeXkaX5UaJJzmT4VEq9A0uVbz1cg8TKmxMukXj2FiykRlT7V3NtdegJcP5L3sezdwcUHnM0PB2cfwlx8CXP2AVcfCPC9tawS2Luh/H979x7bVP3+Afx9enodbOMyxpjapRi+vyF4Y9OE6zDozEZQQRRRLsFLfssAmYvKUBNuYUNuEp2AM0QiBCXEGxgQGyRMLgpOQMULMe7HCG6O7Ytrx9b2tP38/mjXrewUNtg4ZX2/kpP2fHoOe/ZAz8PTc/o5AP5tUvCP04V/HG7UOlyodbrxj8MVXNy44HSj1umC4hOoa/SgrtGD36qv/mvLOgl94wxtGiQT+vYyoF8vE/pYDEiwGJAYXBIs+tBzi0Hmf1SpWwQukfsfPDxyhNahEFGUc7vdqKmpAQBcvHgRpaWlaGxsxKRJk1S3Hzx4MHbs2IF9+/bBZrNh69atOH78eNjZlCVLliAvLw/JycnIycmB0+nE4cOHMX/+/LA/a/Lkydi6dStmzpwJvV6PqVOnqv7MxYsXY8KECbj99tvx1FNPwev1Yu/evXj11VdhtVphNBrxzjvvIC8vD7/88guWL18etn9aWhokScKXX36J3NxcWCwWnD59Gvv370d2djaSk5Px/fff48KFCxg6dOj1pLNLsBki6omMvYAB/wksbfgUBd/s2YPcnBwYfE3BJumypqmxNvhdpeDi9wZuMOtuCJyBuhpJhmRORF9LX/S19EG6pS9g7hM4CxXfB0huXfebEuGQeqNWsaBasaC2CbjY5EH9JQ/+2xg4q/TfpuBjowdOtxc+f2vz1BkGWUKCuaVJats0BRqmXkYdKmskeE9VIzHOhN5mPXqb9OhlCjz2NulhNujYUJGqM0kPIiPBrHUYRBTlvvrqKwwaNAhA4KxOeno6du7cifHjx6tuP2fOHPz++++YNm0aJEnC9OnTkZ+fj71794a2mT17NlwuF9566y28/PLLSEpKitjoTJ06FX6/HzNnzoROp8OUKVPabTN+/Hjs3LkTy5cvx8qVK5GQkIBx4wI3oR8wYAC2bNmC1157DW+//TZGjBiBNWvW4JFHHgntf8stt2Dp0qUoKirCnDlzMGvWLCxcuBDl5eVYv349HA4H0tLSsHbtWuTk5FxrKruMJK5hTrsNGzZg9erVqK6uxrBhw7B+/XqMHTs24vYHDx5EYWEhTp8+jdTUVLz66qvIy8vr8M9zOBxITExEXV1dxKkHY5GiKNizZw9yc3NhMLS/iWcsY27UdTovQgTOHF2qC07kcAFovNBmYoe6Ns1TXeDSvOshGwFTfGAxBh9NvYPrveE19oZL6oVGYYZTWPCv34h/vWbUK0bUeoyo9+hRrxhQ59LhgkuHiy4/HM0KvP6umbpT1knoZZQRbzagl0kOa5bijHrEGWXEGWVYjDIshpbn+tBYnCH42DIeXDfpo7fJqq+vR1JSEhoaGpCQkHD1HWJIS21qKIpHxeRyPHD/PVqHFBV4/FXHvETWmdy4XC5UVlbCZrPBbO7ZH0D4/X44HA4kJCREnG2uJ7vS33VX1qZOnxnasWMHCgoKsGHDBowePRrvvfcecnJy8Ouvv8JqtbbbvrKyErm5uXjhhRewbds2HD58GPn5+RgwYAAef/zx6wqeiLqZJLVejnfZWSZViivwfaXmi0Bz8LGj68IP+DxAU31gUaEH0Du4dOhOLnozRKIFwhAHn2yBT7bAozPDI5ngksxohhlNwohLwohGnwH/OFyQLH3Q6DfA6ZXhVGQ4vDIavDq4hBFuYYDbbYTLbUATDLgojHDDADcM8EAPgWsrVjoJMOllmA06mA2B5shskGEyyDDrdaHHtq+ZDbrQPia9DKNeF1hkXetzvQ6my9bbvm6SA/sZZAmyTorahizancIQjB5xl9ZhEBHRNeh0M7Ru3To899xzeP755wEA69evx759+7Bx40aUlJS0237Tpk2wWq1Yv349AGDo0KH44YcfsGbNGjZDRD2NwQwYUoD4Tt50UojAzWhdDsDTCLid4csVxxyAO/hcaQq/8Z7XBcnrgtR8EToABgBX/RxRuWxdCu7YAT5Jhk8ywivp4YUBiqSHAj08MMAj9HALPTxChkvIcPmD6whsowgZXo8MxaOHFzJ8kKFAhlcEHyHDC33o+b+QoYjAtkpwey908EMX2F8E1lvGfcH9fMHX/dDBKwLjPujgl3TQyXrIsh46XeC5XtbD73F19G8xZtWmPgijPvY+tSUi6gk61Qx5PB5UVFSgqKgobDw7OxtHjhxR3efo0aPIzs4OG3v44YexefNmKIrCU8VEFDgDZU4M3RvpuggBKM2tjZGnSf35Zes+VyPO/d+fsA5Khs7vCZzl8roAr7vNY3P4utIMoPUSPFn4IItmGDsSZzT/31kA8AYWh1ugC/5WerS0kfxgj4joZtWpZqiurg4+nw8DB4bfk2TgwIGhmTEuV1NTo7q91+tFXV1d6Etkbbndbrjd7tC6w+EAELimVFEu/9g2drXkgjlpj7lRFzN5kQyAMTGwdHC2VUVRcMpuR/JDD3X8QxohAL8SaJx8HsCnAD538NEDKfS8dSy0eD2QfJ7A/l53YKIKnxJ49Cutz30KJJWxdtv5fYDfC0n4g+vesHH4vYFLEVXGJXTNd6p6uki16dZbrT3/PdUJMXOc6STmJbLO5EZRFAgh4Pf7Qzf77Klavtbf8vvGGr/fDyEEFEWBLIffxL0r30fXNJvc5deVCyGueK252vZq4y1KSkqwdOnSduMHDhxAXFznbm4VC+x2u9YhRC3mRh3zEln358YQXDpxTwxdcOmu+T+FgAQ/JOEPPSL4vPlSE4D/7aYffHNhbeocHmfUMS+RdSQ3er0eKSkpaGxsvOJNPnsSp/M6Jye6SXk8HjQ3N6O8vBxerzfstaampgh7dV6nSmtSUhJkWW53Fqi2trbd2Z8WKSkpqtvr9fqIM8MtWrQo7AZMDocDt912Gx544AHOJteGoiiw2+14qDOfZMcI5kYd8xIZc6Ouvr4ebIYCWJs6hu8ldcxLZJ3JjdvtRlVVFeLi4nr8hxBCCDidTsTHx8fkBDdNTU2wWCzIysqCyRR+8/dAbeoanWqGjEYjMjIyYLfbMXny5NC43W7Ho48+qrrPyJEjsXv37rCxr7/+GpmZmRH/wZtMpna/NAAYDAYeQFQwL5ExN+qYl8iYm3DMRSvWps5hXtQxL5F1JDeyLIc+mB8wYACMRmOPbRT8fj88Hg/cbndMTa0thIDH48GFCxcgyzLi4uLa/f5d+R7q9EUXhYWFmDlzJjIzMzFy5EiUlZWhqqoqdN+gRYsW4fz58/jwww8BAHl5eSgtLUVhYSFeeOEFHD16FJs3b8ZHH33UZb8EEREREfV8Op0ONpsN1dXV+Pvvv7UOp1sJIdDc3AyLxdJjG74riYuLg9Vq7fZGsNPN0LRp01BfX49ly5ahuroaw4cPx549e5CWlgYAqK6uRlVVVWh7m82GPXv24KWXXsK7776L1NRUvP3225xWm4iIiIg6zWg0wmq1wuv1wufzaR1Ot1EUBeXl5Rg3blzMnU2UZRl6vf6GNIHX9HXc/Px85Ofnq762ZcuWdmNZWVn48ccfr+VHERERERGFkSSpx19yKMsyvF4vzGZzj/49tRY7FyASERERERG1wWaIiIiIiIhiEpshIiIiIiKKSd11C78u1XKTVqfTyWsm21AUBU1NTXA4HMzLZZgbdcxLZMyNupab/bUch6kVa5M6vpfUMS+RMTfqmJfIurI23RTNUMuNlWw2m8aREBHFpvr6eiQmJmodRlRhbSIi0lZX1Kabohnq168fAKCqqorFuI2Wu5+fO3cOCQkJWocTVZgbdcxLZMyNuoaGBlit1tBxmFqxNqnje0kd8xIZc6OOeYmsK2vTTdEMtdxsKTExkf8YVCQkJDAvETA36piXyJgbdbF09/OOYm26Mr6X1DEvkTE36piXyLqiNrG6ERERERFRTGIzREREREREMemmaIZMJhMWL14Mk8mkdShRhXmJjLlRx7xExtyoY14iY27UMS/qmJfImBt1zEtkXZkbSXC+VCIiIiIiikE3xZkhIiIiIiKirsZmiIiIiIiIYhKbISIiIiIiiklshoiIiIiIKCZFfTO0YcMG2Gw2mM1mZGRk4Ntvv9U6JM2VlJTgvvvuQ3x8PJKTk/HYY4/hjz/+0DqsqFNSUgJJklBQUKB1KFHh/PnzmDFjBvr374+4uDjcc889qKio0DosTXm9Xrzxxhuw2WywWCwYPHgwli1bBr/fr3VoN1x5eTkmTZqE1NRUSJKEzz//POx1IQSWLFmC1NRUWCwWjB8/HqdPn9Ym2CjA2tQea1PHsDaFY21qj7Wp1Y2oTVHdDO3YsQMFBQV4/fXXceLECYwdOxY5OTmoqqrSOjRNHTx4EHPnzsV3330Hu90Or9eL7OxsXLp0SevQosbx48dRVlaGu+66S+tQosLFixcxevRoGAwG7N27F7/++ivWrl2LPn36aB2apt58801s2rQJpaWl+O2337Bq1SqsXr0a77zzjtah3XCXLl3C3XffjdLSUtXXV61ahXXr1qG0tBTHjx9HSkoKHnroITidzhscqfZYm9SxNl0da1M41iZ1rE2tbkhtElHs/vvvF3l5eWFj6enpoqioSKOIolNtba0AIA4ePKh1KFHB6XSKIUOGCLvdLrKyssSCBQu0DklzCxcuFGPGjNE6jKgzceJE8eyzz4aNTZkyRcyYMUOjiKIDAPHZZ5+F1v1+v0hJSRErV64MjblcLpGYmCg2bdqkQYTaYm3qGNamcKxN7bE2qWNtUtddtSlqzwx5PB5UVFQgOzs7bDw7OxtHjhzRKKro1NDQAADo16+fxpFEh7lz52LixIl48MEHtQ4lauzatQuZmZl44oknkJycjHvvvRfvv/++1mFpbsyYMdi/fz/OnDkDADh16hQOHTqE3NxcjSOLLpWVlaipqQk7HptMJmRlZcXc8Zi1qeNYm8KxNrXH2qSOtaljuqo26bsjuK5QV1cHn8+HgQMHho0PHDgQNTU1GkUVfYQQKCwsxJgxYzB8+HCtw9Hcxx9/jB9//BHHjx/XOpSo8tdff2Hjxo0oLCzEa6+9hmPHjuHFF1+EyWTCrFmztA5PMwsXLkRDQwPS09MhyzJ8Ph9WrFiB6dOnax1aVGk55qodj8+ePatFSJphbeoY1qZwrE3qWJvUsTZ1TFfVpqhthlpIkhS2LoRoNxbL5s2bh59++gmHDh3SOhTNnTt3DgsWLMDXX38Ns9msdThRxe/3IzMzE8XFxQCAe++9F6dPn8bGjRtjuuDs2LED27Ztw/bt2zFs2DCcPHkSBQUFSE1NxezZs7UOL+rweNyKubgy1qZWrE2RsTapY23qnOs9HkdtM5SUlARZltt90lZbW9uuA4xV8+fPx65du1BeXo5bb71V63A0V1FRgdraWmRkZITGfD4fysvLUVpaCrfbDVmWNYxQO4MGDcIdd9wRNjZ06FB88sknGkUUHV555RUUFRXhqaeeAgDceeedOHv2LEpKSlhw2khJSQEQ+BRu0KBBofFYPB6zNl0da1M41qbIWJvUsTZ1TFfVpqj9zpDRaERGRgbsdnvYuN1ux6hRozSKKjoIITBv3jx8+umn+Oabb2Cz2bQOKSpMmDABP//8M06ePBlaMjMz8cwzz+DkyZMxW2wAYPTo0e2muD1z5gzS0tI0iig6NDU1QacLPwzKshyT05deic1mQ0pKStjx2OPx4ODBgzF3PGZtioy1SR1rU2SsTepYmzqmq2pT1J4ZAoDCwkLMnDkTmZmZGDlyJMrKylBVVYW8vDytQ9PU3LlzsX37dnzxxReIj48PfUKZmJgIi8WicXTaiY+Pb3dteq9evdC/f/+Yv2b9pZdewqhRo1BcXIwnn3wSx44dQ1lZGcrKyrQOTVOTJk3CihUrYLVaMWzYMJw4cQLr1q3Ds88+q3VoN1xjYyP+/PPP0HplZSVOnjyJfv36wWq1oqCgAMXFxRgyZAiGDBmC4uJixMXF4emnn9Ywam2wNqljbVLH2hQZa5M61qZWN6Q2dcVUd93p3XffFWlpacJoNIoRI0Zwik4RmFpQbfnggw+0Di3qcPrSVrt37xbDhw8XJpNJpKeni7KyMq1D0pzD4RALFiwQVqtVmM1mMXjwYPH6668Lt9utdWg33IEDB1SPK7NnzxZCBKYwXbx4sUhJSREmk0mMGzdO/Pzzz9oGrSHWpvZYmzqOtakVa1N7rE2tbkRtkoQQ4rpaNiIiIiIioptQ1H5niIiIiIiIqDuxGSIiIiIiopjEZoiIiIiIiGISmyEiIiIiIopJbIaIiIiIiCgmsRkiIiIiIqKYxGaIiIiIiIhiEpshIiIiIiKKSWyGiIiIiIgoJrEZIiIiIiKimMRmiIiIiIiIYhKbISIiIiIiikn/D5jvRprbUGlkAAAAAElFTkSuQmCC", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(10, 3), sharey=True)\n", - "\n", - "for key, value in lambdas.items():\n", - " lower, median, upper = value\n", - " ax[0].plot(t, S0(median * t), label=f\"{key}\")\n", - "\n", - "ax[0].grid(True)\n", - "ax[0].set_xlim(0, 10)\n", - "ax[0].legend()\n", - "ax[0].set_title(\"Probability Not Being Adopted by Time $d_k$\")\n", - "\n", - "for key, value in lambdas.items():\n", - " lower, median, upper = value\n", - " ax[1].plot(t, cdf(median * t), label=f\"{key}\")\n", - "\n", - "ax[1].grid(True)\n", - "ax[1].set_xlim(0, 10)\n", - "ax[1].legend()\n", - "ax[1].set_title(\"Probability of Being Adopted by Time $d_k$\");" - ] - }, - { - "cell_type": "code", - "execution_count": 31, + "execution_count": 82, "metadata": {}, "outputs": [], "source": [ @@ -1034,19 +976,19 @@ "black_cats_preds = np.quantile(black_cats, [0.025, 0.5, 0.975])\n", "\n", "lambdas = {\n", - " \"Other cats\": other_cats_preds,\n", - " \"Black cats\": black_cats_preds\n", + " \"Other cats\": 1 / other_cats_preds,\n", + " \"Black cats\": 1 / black_cats_preds\n", " }" ] }, { "cell_type": "code", - "execution_count": 32, + "execution_count": 83, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0MAAAEpCAYAAABRMPn6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACRLElEQVR4nOzdd3gU1dfA8e9sS+8dEkjogVBDR5pIVRQRCyoIdrEhor6KBbArFhTRnw1UEBW79Ij0DlJDLyEBEtJ72TbvHwPBkAQSCGzK+TzPPsneaWdvZvfuyb1zR1FVVUUIIYQQQggh6hidowMQQgghhBBCCEeQZEgIIYQQQghRJ0kyJIQQQgghhKiTJBkSQgghhBBC1EmSDAkhhBBCCCHqJEmGhBBCCCGEEHWSJENCCCGEEEKIOkmSISGEEEIIIUSdJMmQEEIIIYQQok6SZEgIIYQQQghRJ0kyJIQQQgghhKiTamwyNHv2bBRFKX4YDAZCQ0MZO3YsJ0+erNJjTZ48GUVRSE1Nver7O/s64+Liyi0ra53169czefJkMjMzqyTm84/t7OzM8ePHSy3v06cPUVFRld5vZeM9/++vKAoBAQH06dOHBQsWVPr45+/3v3XpCB999BGKolS6Ls+eW1falTq/Khp/Vb8ny3L++VXeY+XKldXmvKkou92Ou7s7Tz/9tKNDqTOkzXJMm1UZP/74I61atcLFxQVFUdixY0eZ60n7U7ar0f5cyfNE2p8rr7q2PTU2GTpr1qxZbNiwgZiYGB544AHmzZtHz549ycvLc3RoVeL6669nw4YNhISEVGqd9evXM2XKlCvWsBQVFfHiiy9W2f4uNd6zf//169fz+eefo9frGTp0KH/99dclxVGR+r4avv76awBiY2PZtGmTQ2Mpy5U+v6qDDRs2lHgMGTIEFxeXUuUdOnSoNudNRcXGxpKXl0enTp0cHUqdI22WY9qsi0lJSWHUqFE0btyYJUuWsGHDBpo1a3bBbaT9ufocfZ5cLbW1/amubY/B0QFcrqioKDp27AhA3759sdlsvPrqq/z+++/cddddZW6Tn5+Pq6vr1QzzkgUEBBAQEHDZ61S1QYMG8f333zNx4kTatm17VY/9X//9+5+Ny8fHh3nz5jF06NBK788RdXm+rVu3snPnTq6//noWLlzIV199RZcuXRwaU13UtWvXEs8DAgLQ6XSlyv+7vKbYvHkzQLVrkOoCabOqx+fs+Q4ePIjFYuHuu++md+/eFdpG2h9xpdTW9qe6tj01vmfofGdPlLNDuM52Z/7777+MGDECHx8fGjduDMDatWvp168fHh4euLq60r17dxYuXFjuvhMSEhg+fDienp54eXlx9913k5KSUrz88OHDjB07lqZNm+Lq6kr9+vUZOnQou3fvvqT9QcW6zc9fZ/LkyTzzzDMAREREFHenvvrqqyiKwrx580rt49tvv0VRFLZs2VLucf7r2Wefxc/Pj+eee+6i616snsuLd+XKlRWK5b+cnZ0xmUwYjcYS5YcOHeLOO+8kMDAQJycnIiMj+eSTT0ptX1ZdKopCbGwsI0eOxMvLi6CgIO69916ysrJKbf/HH3/Qpk0bnJycaNSoEdOnT6/00IGvvvoKgLfeeovu3bvzww8/kJ+fX2q9hQsX0q5dO5ycnIiIiGDatGnl7rMi5/rZOLdv337B8/Jif6+K1nVl4i/Pxd5Da9asqbJz/kLKep+erc9du3Zx66234uXlha+vLxMmTMBqtXLgwAEGDRqEh4cH4eHhvPPOO6X2W9G6vJAvvviC1q1b4+zsTFRUFEuXLmXz5s0lPguF40ibdeXbrIvV25gxY7jmmmsAuP3221EUhT59+lx0v+eT9qdsVdX+VOS7grQ/murQ/tSotketoWbNmqUC6pYtW0qUT58+XQXUzz//XFVVVX3llVdUQG3YsKH63HPPqTExMervv/+urly5UjUajWp0dLT6448/qr///rs6YMAAVVEU9Ycffiixz//u45lnnlGXLl2qvv/++6qbm5vavn171Ww2q6qqqqtWrVKffvpp9eeff1ZXrVql/vbbb+qwYcNUFxcXdf/+/ZXe339f57Fjx8otO/95QkKC+vjjj6uA+uuvv6obNmxQN2zYoGZlZant27dXe/ToUao+O3XqpHbq1KlS9X62rpcvX168vHfv3mqrVq2Kn1ekni8U78Xi2Lhxo2qxWFSz2awmJCSoTzzxhKrT6dQlS5YUrxsbG6t6eXmprVu3Vr/99lt12bJl6tNPP63qdDp18uTJZe73bF2e/Vs1b95cffnll9WYmBj1/fffV52cnNSxY8eW2Hbx4sWqTqdT+/Tpo/7222/q/Pnz1S5duqjh4eFqRd9q+fn5qpeXV/Hf4ssvv1QBdfbs2SXW+/vvv1W9Xq9ec8016q+//qrOnz9f7dSpk9qgQYNSx6rouV7R8/JCf6+K1nVl4i9LZd5Dl3vOn3XPPfeobm5uZS4r633633Pn1VdfVWNiYtRnn31WBdTHHntMbdGihfrRRx+pMTEx6tixY1VA/eWXX4q3r8x5W54nn3xSdXZ2Vl988UX1n3/+UWfOnKkGBQWpYWFhav/+/Sv82sXlkzbLMW1WRert8OHD6ieffKIC6htvvKFu2LBBjY2NLXef0v44pv252HcFaX+OlYrRUe1PTWt7anwydPbDKCcnR12wYIEaEBCgenh4qElJSaqqnjshXn755RLbd+3aVQ0MDFRzcnKKy6xWqxoVFaWGhoaqdru9uPzsPp566qkS+5g7d64KqHPmzCkzRqvVqprNZrVp06Yltq3M/i6lYVFVVX333XdLlf133e3btxeXbd68WQXUb775pszXUdb2W7ZsUYuKitRGjRqpHTt2LK6v85OhitZzefFeLI7zH05OTurMmTNLrDtw4EA1NDS0VHL12GOPqc7Ozmp6enqp/Z7fGL3zzjslth03bpzq7Oxc4jzp1KmTGhYWphYVFRWX5eTkqH5+fhVujL799lsVUD/77LPi7d3d3dWePXuWWK9Lly5qvXr11IKCguKy7Oxs1dfXt9SxKvo3qMx5Wd7fq6J1XZn4y3Ip76FLPefPutTG6L333iuxbrt27Yob87MsFosaEBCgDh8+vLisMudtWX7++WcVKPVF+Y033lAB9YUXXlBVVVXtdrvq7u6uJicnX3B/4vJIm+WYNqui9bZixQoVUOfPn3/RfUr747j250LfFaT9OVYqRke0PxVte1S1+rQ/NX6YXNeuXTEajXh4eHDDDTcQHBzM4sWLCQoKKrHeLbfcUvx7Xl4emzZtYsSIEbi7uxeX6/V6Ro0axYkTJzhw4ECpY50/nvu2227DYDCwYsUKAKxWK2+88QYtW7bEZDJhMBgwmUwcOnSIffv2VXp/V8LIkSMJDAws0dX58ccfExAQwO23316pfZlMJl577TW2bt3KTz/9VGr5pdZzZXz77bds2bKFLVu2sHjxYu655x4effRRZsyYAUBhYSHLly/n5ptvxtXVFavVWvwYMmQIhYWFbNy48aLHufHGG0s8b9OmDYWFhSQnJxe/1q1btzJs2DBMJlPxeu7u7pUaO/7VV1/h4uLCHXfcUbz9rbfeypo1azh06FDxsbZs2cLw4cNxdnYu3tbDw6PUsS7lb3Cp52VF67oy8V9MRWKtynP+Utxwww0lnkdGRqIoCoMHDy4uMxgMNGnSpHioVFWct6+++iqdOnUq9RpbtmwJUHytw6FDh/Dx8akxY85rOmmzKudy3r9Xug2S9kfan4vFWhfbn4q2PVB92p8anwyd/TDavn07p06dYteuXfTo0aPUev+dZSMjIwNVVcuceaNevXoApKWllVoWHBxc4rnBYMDPz6943QkTJvDSSy8xbNgw/vrrLzZt2sSWLVto27YtBQUFld7fleDk5MRDDz3E999/T2ZmJikpKfz000/cf//9ODk5VXp/d9xxBx06dGDSpElYLJYSyy61nisjMjKSjh070rFjRwYNGsT//vc/BgwYwLPPPktmZiZpaWlYrVY+/vhjjEZjiceQIUMAKjQ9pp+fX4nnZ+vq7N/17Gs9/wsNUGZZWQ4fPszq1au5/vrrUVWVzMxMMjMzGTFiBHBuhp+MjAzsdnup8wdKn1OX8je41POyonVdmfgvpiKxVvU5X1m+vr4lnptMJlxdXUs0xGfLCwsLgYrXZXmSkpLYuXMnd955Z6llJ06cAM5dwPrvv/8WN04pKSn06tWLF154AVVVL/EViwuRNqtyLuf9e6XbIGl/SpL2R9qfyrQ9UH3anxo/m9zZD6OL+e8FhD4+Puh0OhITE0utd+rUKQD8/f1LLUtKSqJ+/frFz61WK2lpacUfVHPmzGH06NG88cYbJbZLTU3F29u70vu7Uh555BHeeustvv76awoLC7FarTz88MOXtC9FUXj77bfp378/n3/+eYlll1rPl6tNmzYsXbqUgwcPEhUVVfxfqEcffbTM9SMiIi77mD4+PiiKwunTp0stS0pKqtA+vv76a1RV5eeff+bnn38utfybb77htddeKz5WWfs9v+xS/gaXel76+PhUqK6dnZ0rHP/FVDTWqjznr4aK1mV5zjY6ZX0J+f777wkODiY0NBSA7du307FjR3bu3Mntt9/Oiy++yN13310Fr0KURdqsyrvU968j2iBpf0rGJe1P3Wp/KtP2QPVpf2p8MnQp3Nzc6NKlC7/++ivTpk3DxcUF0G4GNWfOHEJDQ8u8v8DcuXOJjo4ufv7TTz9htVqLZ51RFKVUpr9w4UJOnjxJkyZNKr2/y3H+f47+KyQkhFtvvZWZM2diNpsZOnQoDRo0uORjXXfddfTv35+pU6cSFhZWXF6Zer5QvJV19kZ5AQEBuLq60rdvX7Zv306bNm1KDCGoSm5ubnTs2JHff/+dadOmFR8nNze3Qjfhs9lsfPPNNzRu3Jgvv/yy1PIFCxbw3nvvsXjxYm644QY6d+7Mr7/+yrvvvlv8H56cnJxS97e4lHO9IudlWX+vytR1ReO/mIq+h6r6nL/SLve8PTvkYM+ePSWGKvz888+sX7++xNCJ7du306hRI4YPH86cOXPo1q1b1bwIUWWkzbq09++l1tvlkPanZFxV3f6Ud55I+1N1Lue8rUzbA9Wn/amTyRDAm2++Sf/+/enbty8TJ07EZDIxc+ZM9uzZw7x588qcivLXX3/FYDDQv39/YmNjeemll2jbti233XYboI3NnD17Ni1atKBNmzZs27aNd999t0QWXJn9XY7WrVsDMH36dO655x6MRiPNmzfHw8MDgCeffLL43gGzZs267OO9/fbbREdHk5ycTKtWrYrLK1rPF4u3PHv27MFqtQJa1+6vv/5KTEwMN998c/F/LqZPn84111xDz549eeSRRwgPDycnJ4fDhw/z119/8c8//1z26weYOnUq119/PQMHDuTJJ5/EZrPx7rvv4u7uTnp6+gW3Xbx4MadOneLtt98u84tFVFQUM2bM4KuvvuKGG27g1VdfZdCgQfTv35+nn34am83G22+/jZubW6ljVfZcr8h5Wd7fq6J1XZn4L6Qy76GqPuevtMs5bxs0aECnTp344IMPCAgIoE2bNqxevZrp06cDJYcpbN++nR07djBt2jRJhKoxabMu7f17KfVWUdL+XP3250LnibQ/VedSz9vKtD1QjdofR8zaUBXKm6b0fGdn1EhJSSm1bM2aNeq1116rurm5qS4uLmrXrl3Vv/76q9x9bNu2TR06dKjq7u6uenh4qCNHjlRPnz5dvF5GRoZ63333qYGBgaqrq6t6zTXXqGvWrFF79+6t9u7du9L7++/rrOzMPKqqqs8//7xar149VafTqYC6YsWKEsvDw8PVyMjIC9bf+S5U73feeacKlJhNTlUrXs8Xi7esOP778PLyUtu1a6e+//77amFhYYn1jx07pt57771q/fr1VaPRqAYEBKjdu3dXX3vttTL3e/5sPuefP+XV+W+//aa2bt1aNZlMaoMGDdS33npLfeKJJ1QfH59yX4uqquqwYcNUk8l0wRlV7rjjDtVgMBTPOvXnn3+qbdq0KXGss/GeryJ/g8qcl6pa/t+ronVdmfjPV9lYz7qUc/6sS53N5/xzp7z9nD8To6pWvC7LcuzYMXXQoEGqu7u76u3trQ4dOlT96quvVEBduHChqqqqGh8fr9arV09du3atGhISosbFxV10v+LSSJvlmDarovV2ubPJSftz9dqfC50n0v6UjNER7U9F2h5VrV7tT41NhsTl2blzpwqon3zyiaNDqdXMZrPasmXLajmv/vku9CWsNpBzvrQ//vhDHThwoKqqqjpz5ky1ffv2an5+voOjEqI0ef9WnrQ/1Yecv6VVp/anzg6Tq6uOHDnC8ePHeeGFFwgJCWHMmDGODqlWue++++jfvz8hISEkJSXx2WefsW/fvuIuYnH1yTlfvrNjwkG7yHfr1q3cf//9zJ0718GRCaGR92/FSftT/cj5W77q1P5IMlTHvPrqq3z33XdERkYyf/58XF1dHR1SrZKTk8PEiRNJSUnBaDTSoUMHFi1axHXXXefo0OosOefL98orr5R4/tVXXzkoEiHKJu/fipP2p/qR87d81an9UVRVbiYhhBBCCCGEqHtq/E1XhRBCCCGEEOJSSDIkhBBCCCGEqJMkGRJCCCGEEELUSTViAgW73c6pU6fw8PC4rBukCSGEqBxVVcnJyaFevXrodPL/s/+StkkIIRyjKtumGpEMnTp1irCwMEeHIYQQdVZCQgKhoaGODqNakbZJCCEcqyraphqRDHl4eABw7NgxfH19HRxN9WGxWFi2bBkDBgzAaDQ6OpxqReqmbFIv5ZO6KVt6ejoRERHFn8PiHGmbyibvpbJJvZRP6qZsUi/lq8q2qUYkQ2eHH3h4eODp6engaKoPi8WCq6srnp6e8iY5j9RN2aReyid1UzaLxQIgw8DKIG1T2eS9VDapl/JJ3ZRN6qV8Vdk2yQBwIYQQQgghRJ0kyZAQQgghhBCiTpJkSAghhBBCCFEnVfqaodWrV/Puu++ybds2EhMT+e233xg2bNgFt1m1ahUTJkwgNjaWevXq8eyzz/Lwww9fasxCiKvMbrdjNpsdHcYVZbFYMBgMFBYWYrPZHB3OVWM0GtHr9Y4OQwghhHCISidDeXl5tG3blrFjx3LLLbdcdP1jx44xZMgQHnjgAebMmcO6desYN24cAQEBFdpeCOFYZrOZY8eOYbfbHR3KFaWqKsHBwSQkJNS5yQK8vb0JDg6uc69bCCGEqHQyNHjwYAYPHlzh9T/77DMaNGjAhx9+CEBkZCRbt25l2rRpkgwJUc2pqkpiYiJ6vZ6wsLBafdNNu91Obm4u7u7utfp1/peqquTn55OcnAxASEiIgyMSQgghrq4rPrX2hg0bGDBgQImygQMH8tVXX2GxWGSqQCGqMavVSn5+PvXq1cPV1dXR4VxRZ4cCOjs715lkCMDFxQWA5ORkAgMDZcicEEKIOuWKJ0NJSUkEBQWVKAsKCsJqtZKamlrmfyKLioooKioqfp6dnQ2AuaioeF5xcW6OdamT0qRuylbZeikqKkJVVQwGQ50YJnf2Z21/redzdnZGVVUKCgpwcnIqsUzeQ+eU1zZZLBapp/+Qz9+ySb2UT+qmbJdSL6qqYrWrWGx2LDYVq82Oxa5is6tYbVq5zX5unXO/q9jsdm2d4vXt2FTtd5udM8vtqHYrqs0CNit2mwXFZsFus4DNAnYL2K1nfreC3Yqi2sCuPRS7FVStTLHbzvxuRzmznqJaUew2dKpN2061o8OmraPaUVQbOuzk5xdUWT1flZuunj8O/eyXjvLGp7/55ptMmTKlVPmy3+fgXb951QdYw8XExDg6hGpL6qZsFa0Xg8FAcHAweXl5daaRysnJcXQIV53ZbKagoIBVq1ZhtVpLLMvPz3dQVNVPeW3TihUran3P6aWQz9+ySb2Ur6bXjV0Fsx3MtjM/7WCxgUUFq13BYger/exzzj23g0VVSpVZVbDbVL6KXYjBbkGvnnnYzRiwYFTN6O0WjFgwqGaMqva7ExacseCkmDFhxXjmYcKCSbFhwvKfMismxYprcZntzHracgM2jNjO/a5UjwmGss0qY6toX1c8GQoODiYpKalEWXJyMgaDAT8/vzK3ef7555kwYULx8+zsbMLCwgizHqXnkKeuaLw1icViISYmhv79+8tww/NI3ZStsvVSWFhIQkIC7u7uODs7X4UIHUdVVXJycvDw8KhzEwkUFhbi4uJCr169Sv2d09LSHBRV9VNe29S3b99y27O6SD5/yyb1Uj5H1I3VZifPbCOvyEpukZXcIpv2s/Dscyt5Z8ryzDYKLTbyz/wsOPPIN9soNGs/VUsBTrY83JUC3CjAQynAnQLcKMRVKcKFMw+lCNfi3824UoQzRbieKT/7u7a++cLJhwJUg5HNdhTsigGboseuGEo8VEWPXdGjKgZURYeqM5x5rkfVGVB1BvjPcxQ9qk4Pih50Wrn2u654PXR6svMtwOdVEv8VT4a6devGX3/9VaJs2bJldOzYsdwT3snJqdRQDYD6ScsxGgxQx76oXIzRaJQP1nJI3ZStovVis9lQFAWdTlfrr6M5OzROURS+/fZbxo8fT2ZmpmODukp0Oh2KopR5Xsj755zy2ib5nCmb1EvZpF7KV9m6KbTYyMg3k5lvITPfQlaB9ntGvoXMAjNZZ8qzCy3knE1yzvwssJxLMkxY8CIPTyUPb3LxOvPTW8nDS8mjPlpi467k404h7mcTHaUQD/JxpwC9UYUr/Ge164zYdSbseidUvROqwQn0TmBw1h5GZxSjM4rRBd2Z33VGFxSDSVtPbwK9EQxO2s9SZabzHkbtoTOC3nDmpxF0hv+UG9Hp9Oi4SsPNzkhLS4OHHJQM5ebmcvjw4eLnx44dY8eOHfj6+tKgQQOef/55Tp48ybfffgvAww8/zIwZM5gwYQIPPPAAGzZs4KuvvmLevHmVDjZEPU3K4a0ENO1U6W2FEHVLQkICkydPZvHixcXXJw4bNoyXX365xH/xw8PDGT9+PE888YQDo708ffr0oV27dsWzdgohRE2jqpBTaCUzs4iUnCJScotILf5pJrPATHqe+UzSoz2KrKWv73ShEH8lC3+ytZ9KFo3IOZPgaMmNt5KLlykPLyUXb/JwVYrKiOgSXgMKGF3ByR3F5AYm9zMPVzC6gcnlzM+zD/dzvxtdwehy5ndtPYtiImblGvoPHorRWZvptHb/W9IxKp0Mbd26lb59+xY/Pztk4J577mH27NkkJiYSHx9fvDwiIoJFixbx1FNP8cknn1CvXj0++uijS55WO2n9PEmGhBAXdPToUbp160azZs2YN28eERERxMbG8swzz7B48WI2btyIr6/vVY9LZtAUQtQ1Vpud0zlFnM4uLE5uUnK0R3JOEck5Wnlyth7Lxn/K2IOKF3n4K1kEKFk0+0+S42fIIlCXTaBOK/NVs3DmUhMbBZzcwckDnL3AyQtcvLWHkxc4e5776ewFzt7auv95KEZXbThXVbFYsBjcteSolo/OcKRKJ0N9+vQpngChLLNnzy5V1rt3b/7999/KHqpMgQmLQX1PhsoJIcr16KOPYjKZWLZsWfHU0Q0aNKB9+/Y0btyYSZMm8emnn9KnTx+OHz/OU089xVNPadcj2mznhk4sXbqU8ePHk5CQwDXXXMOsWbNKzIA5a9Ys3nnnHY4dO0Z4eDhPPPEE48aNAyAuLo6IiAh+/PFHZs6cycaNG/n0008ZO7b0JZ+ZmZk8++yz/PHHH2RlZdGkSRPeeustbrjhBtLS0njsscdYs2YN6enpNG7cmBdeeIGRI0cCMGbMGFatWsWqVauYPn06oPXYe3l58dhjj7Fs2TJyc3MJDQ3lhRdeKPP4QghxqYqsNhIzCzmZWcCJjHxOZhSQkFFAQno+JzMLSM4uwnaB743u5BOipNNdSSNEn0ZDfTrhxnRClVQCScPXloZRNV88kP8eQm8CFx9w9QUXX3D1O5PY+ICLn/bTzQ9cA7Rlbr7g5Am6anABjrjqrubwvstWoBoJsp4i5/i/eIRHOzocIeocVVVLjLO+mlyM+gpNbJCens7SpUt5/fXXixOhs4KDg7nrrruKE5Rff/2Vtm3b8uCDD3LfffeVmEkuPz+fadOm8d1336HT6bj77ruZOHEic+fOBeCLL77glVdeYcaMGbRv357t27fzwAMP4Obmxj333FO8n+eee4733nuPWbNmlXm9id1uZ/DgweTk5DBnzhwaN27M3r17i+/3U1hYSHR0NM899xyenp4sXLiQUaNG0ahRI7p06cL06dM5ePAgUVFRTJ06FYCAgACefPJJ9u7dy+LFi/H39+fw4cMUFFTdVKRCiLrBZldJSM8nLi2PExkFnMzIJz49nxMZBZzK0np1ykt1FOwEkUETfRLNnTOIMKQTpksjmFT87Kl4WtMw2cqYsbKsZsbkpvXGuPieSWT8wC1Ae3gEg3uw9tOzvpb4yD/NRQXVqGRom6EDQ9hM4rrvJRkSwgEKLDZavrzUIcfeO3UgrqaLf2QdOnQIVVWJjIwsc3lkZCQZGRmkpKQU32TUw8OD4ODgEtMjWywWPvvsMxo3bgzAY489VpxsALz66qu89957DB8+HNCGBO/du5f//e9/JZKh8ePHF69Tlr///pvNmzezb98+mjVrBkCjRo2Kl9evX5+JEycWP3/88cdZsmQJ8+fPp0uXLnh5eWEymXB1dSU4OLh4vfj4eNq3b0/Hjh0B7dooIYQoT5HVxrHUPA4n53IwKYeDybkcTs4lPi0fs+1C915TCTHk0NE1hdZOp2lqSCTMfooASyLuRYnobWeGrdkoO8kBVJMb2TpfPPyC0XkEg2c98KoPXg3AJxx8GmhD1PQ16murqCFq1FmV1aA/nNqMT9xCUKdJ1i+EqLSL3efsLFdX1+JECCAkJITk5GQAUlJSSEhI4L777uOBBx4oXsdqteLl5VViP2eTkfLs2LGD0NDQ4kTofDabjbfeeosff/yRkydPFt/4083N7YL7feSRR7jlllv4999/GTBgAMOGDaN79+4X3EYIUfvlFFo4kpLHodM5HDydw8HTuRxNyeVkZgH2crp4THqFll4W2rsk0dKYRCPlFCHWk3ibE3HOP4XOkg9mtMf5FB24B2k9N+4B53pvvELBKwx8wrGavFj59yqGDBmCTq6rFFdZjUqGQjteT8Ef7xJgSaQo4V+cGkjvkBBXk4tRz96pAx127Ipo0qQJiqKwd+9ehg0bVmr5/v378fHxwd/f/4L7OX+iA0VRihOps9Nwf/HFF3Tp0qXEemeHt511saTl/KF853vvvff44IMP+PDDD2ndujVubm6MHz8es/nCY+gHDx7M8ePHWbhwIX///Tf9+vXj0UcfZdq0aRfcTghReyRlFbIjIZN/4zPYdSKTIyl5pOSUP8GAh5OeRj4Gunqk0tF4jKa2IwQVHsM5+yhKfjqUew9mRRuu5llPS3S8G4BvIwhoCr5NtEkHDC5gMJW9eR25qbeonmpUMtQiLIh1+o5cZ19P0oYfaCjJkBBXlaIoFRqq5kh+fn7079+fmTNn8tRTT5VINpKSkpg7dy6jR48u7hkymUwlJk2oiKCgIOrXr8/Ro0e56667LiveNm3acOLECQ4ePFhm79CaNWu46aabuPvuuwEtETt06FCJYYDlvYaAgADGjBnDmDFj6NmzJ88884wkQ0LUUtmFFnafyGLb8XS2x2ey+2QWqbll/9PEz81IYx8TLbxtdHY6Tgv7EeqZj+KSfQwl4xhkFpZ9EBffMwnPmaTHJwL8GoNfE22yAqOLNkW0TEQgapDq/a3mPDqdQmqDIRC3Ho/Df4H6jgyVE0KUMmPGDLp3787AgQN57bXXSkytXb9+fV5//fXidcPDw1m9ejW33XYbZrMZT0/PCh1j8uTJPPHEE3h6ejJ48GCKiorYunUrGRkZxbccqIjevXvTq1cvbrnlFt5//32aNGnC/v37URSFQYMG0aRJE3755RfWr1+Pj48P77//PklJSSWSofDwcDZt2kRcXBzu7u74+voyefJkoqOjadWqFUVFRSxYsKDc66iEEDVLkdXG/sQcth3P4N/4DHafyOJ4euluG50Cjf2caRvsRDefbNpwhFDzEVyyj0L6MThyAtQyrgfSm7RrdXwbgU8jCGwBQVHg5n/uPjgGme5Z1A41KhkCCOtyE/nHXsXXkojt5Hb0oR0cHZIQoppp2rQpW7duZfLkydx+++2kpaURHBzMsGHDeOWVV0rcY2jq1Kk89NBDNG3alKKiogr3Et1///24urry7rvv8uyzz+Lm5kbr1q0ZP358peP95ZdfmDhxIiNHjiQvL694am2Al156iWPHjjFw4EBcXV158MEHGTZsGFlZWcXbT5w4kXvuuYeWLVtSUFDAsWPHMJlMPP/888TFxeHi4kLPnj354YcfKh2bEMLxUnOLWH0whU3H0tl1IpPDyblYbKUv8AnxdKJ1sDMdgvR0c0uiReFOnNJi4fQeOHyq7J07eYJvY/CN0JKfwEgIaH7mPjruYPKQiQtErVbjzu5OTUP5R+nAIDaQsvEHgkdIMiSEKK1hw4bMmjXrout17dqVnTt3Yrfbyc7OBigeWvZfw4YNK3WPtTvvvJM777yzzP2Gh4df8J5s/+Xr68vXX39d7rLff//9gts3a9aMDRs2lCh78cUXefHFFyt0fCFE9WK22tkal87f+5JZcyiFQ8m5pdbxdDYQFexKu2AjHQOgg+Eo3hl7tMRn717ISym9Y896Wk+PbyMt+QlooV3fc/bGoSb38q/rEaKWqnHJkMmg41S9QXBqA84H/wT1bRkqJ4QQQogaS1VV4tLyidmbxKoDKWyLz6DQUnL4WtMAV7qEuhAdoifa10JYwX6U5L1a8rMvFoqyS+5U0YN/E214W1AUBLUGz2At4Tmb+Bidr+KrFKJ6qnHJEEBIxxvJ/+MNvM2JqCf/RQmViRSEEEIIUXNkF1pYfTCFf/Yns/5IGklZJSct8HIx0q2hO9eGm+gTbCEwOxaSdsGBPZC8DyznXSOkN2k9PUGttEdAS/AIPHOjUm8t+ZF/HgtRSo1Mhnq2asCK3ztwvbKBjK0/4SvJkBBCCCGqMbtd5XguTIs5yPojGcSeyipxXx+DTqF1PQ96hbvQv4GOlq6Z6BI3QsIm2LQJCjJK7tDoAoEtz/X8+DcFFx8t8XH21q4FkgkOhLioGpkMuTsZOBbUH1I2YNj3B9z0lvy3QwghhBDVzv7EbOZtSWDRrlOk5BqAuOJl9b2d6RHuQd+GJnoFWXAzp8LJFbBtI5z8F6wF53ZkdIGQdmeSn1batNbOHmd6fny05EcmOhCi0mrsuyaww1DylryNZ1Gi9oEhvUNCCCGEqAaSswv5aWsCf+w4VWLyAye9SqeGvvRp5MqABtDAKR+yE7Ten50b4HQsqP+Z0dLFFxp0hbAuENxGu9bnbM+Ps5dMdiBEFaixyVCfqAYsX9SBG/UbyN0+H3dJhoQQQgjhIPlFVv7YeYrf/j3B1uMZxUPg9DqFLuHeDG/hCmnHuSnKgjF7J8SeGQKXfqTkjrwbQFg3aNBFu5mpkzu4+oNbAJhcr/4LE6KWq7HJUKCHM/t8+3Fj1gaI/Q1ueFOGygkhhBDiqrHZVVYeSGb+1hOsPJhcYga4FsEeDGvpxa1NwY9srKe3cPzUMgyHd0Be8rmdKDoIiNSSn7Cu2vTXJlct+XH1lwRIiCusxiZDAP7thpC7chruhUlwchuEdnR0SEIIIYSo5XadyOTHLQks2ZNEWp65uDzI04kbWvpze6SRZi45UJAI+1fCwcUYMuJofHZFvRPUa68NfwvrrA15M7mBm78kQEJcZTU6Gbq2dTjL/+nATfr1FO38BSdJhoQQQghxBWTmm5m9Po4/d5ziaGpecbmbSc+1zf25vZUbPfzzUMzZkHoAtiyBY2vAVgSAqjdxwrMjIW36YghtDwZnSYCEqAZqdDIU4e/GNx59uCl/Pbbdv8KQN2SonBCiQuLi4oiIiGD79u20a9euSvapKAq//fYbw4YNq5L9CSEc73RWAZ+sPMLP206Qb9YmN9CuA/LllihPrm9ow9mSBUUpsHslHFwCGXHnduDVAJoPwtqwD/8etjKkcX3wDJQESIhqokYnQwDeUYPI3fS+NlTuxFYI6+TokIQQDjZmzBi++eab4ue+vr506tSJd955hzZt2jgwsqq1cuVK+vbtS0ZGBt7e3o4OR4ha5XhqHh/9c5i/dp3CbNWuBQrzdWVk+0BubW4ggAywpsKJ/XBwKcSd6wVCb4Lwa6DZYO1GqHojuPjD4b0Q0haMRge+MiHEf9X4ZOjaNg35e0MHhunXY93zKwZJhoQQwKBBg5g1axYASUlJvPjii9xwww3Ex8c7ODIhRHW2LzGbj5YfYtne09jOTAnXNNCdR7sHcWNYATpzOhTkwpEVcGhpyV4g7wZaAtSorzYLnMlNmxDBLQBsdmCvQ16TEKJ8Nf7WxK3re7HOqRcA1l2/gt1+kS2EEHWBk5MTwcHBBAcH065dO5577jkSEhJISUkpc32bzcbjjz9O48aNcXFxoXnz5kyfPr3Uel9//TWtWrXCycmJkJAQHnvssXJjmDp1KkFBQezYsaPcdf788086duyIs7Mz/v7+DB8+vHjZnDlz6NixIx4eHgQHB3PnnXeSnKzNQhUXF0ffvn0B8PHxQVEUxowZA8DPP/9M69atcXFxwc/Pj+uuu468vLxSxxZCnLM1Lp17vt7MkOlrWLwnCZtdpU2oF5/f2pRlt7kzLCgF3YktsPYD+Oke2Pw/LRHSm6BxPxj8Ltz4CbS8EXwjIKQN1O8AHsGg0zv65QkhylHje4YURcGj1QBydkzHoyAJTm7VZmYRQlQ9VQVLvmOObXS95GsCc3NzmTt3Lk2aNMHPz6/Mdex2O/Xq1eOHH34gMDCQ9evX8+CDDxISEsJtt90GwKeffsqECRN46623GDx4MFlZWaxbt67UvlRVZfz48fz++++sXbuWpk2blnnMhQsXMnz4cCZNmsR3332H2Wxm4cKFxcvNZjOvvvoqzZs3Jzk5maeeeooxY8awaNEiwsLC+OWXX7jllls4cOAAnp6euLi4kJiYyMiRI3nnnXe4+eabycnJYc2aNaiqekl1J0Rtt3J/Mp+sPMyWuIzisi4RvjzezZ9r/LLBfBpO7oXtcyBp17kNvRtCs0HneoH0Ji3x8QiRm6EKUYPU+GQI4Nqohvz9bwdu1q/Dvuc3dJIMCXFlWPLhjXqOOfYLp7QhJxW0YMEC3N3dAcjLyyMkJIQFCxag05XdIW40Gnn++efx9PREp9MRERHB+vXr+emnn4qToddee42nn36aJ598sni7Tp1KDs21Wq2MHj2arVu3sm7dOkJDQ8uN8fXXX+eOO+5gypQpxWVt27Yt/v3ee+8t/r1Ro0Z89NFHdO7cmdzcXNzd3fH19QUgMDCw+JqhI0eOYLVaGT58OA0bNgSgdevWF60vIeoSu93Owt1JfLbqCLGnsgHtfy19mgXweDc/OrhngiUJTh2CHXO023cA6AwQ0VtLggJaaBs5e4JHPXD1g3I+X4QQ1VetSIa6NPJlgr4HN7MO6+5fMA18XT6QhKjj+vbty6effgpAeno6M2fOZPDgwWzevLk4STjf119/zffff8/x48cpKCjAbDYXzzSXnJzMqVOn6Nev3wWP+9RTT+Hk5MTGjRvx9/e/4Lo7duzggQceKHf59u3bmTx5Mjt27CA9PR37mWHA8fHxtGzZssxt2rZtS79+/WjdujUDBw5kwIABjBgxAh8fnwvGIkRdYLXamf/vCb5Yc5SjKdrQUb1OYWDLIB7v5kOkUzpYTsHpY7BjLiRs1DZU9NC0P7S+DdwDtRulugdqvUBO7g58RUKIy1UrkiGjXodTs35kH/gYz/zTcGKLdidnIUTVMrpqPTSOOnYluLm50aRJk+Ln0dHReHl58cUXX/Daa6+VWv+nn35i0qRJTJs2je7du+Ph4cG7777Lpk2bAHBxcanQcfv378+8efNYunQpd9111wXXvdA+8/LyGDBgAAMGDGDOnDkEBAQQHx/PwIEDMZvN5W6n1+uJiYlh/fr1LFu2jI8//phJkyaxadMmIiIiKvQahKiNVh5I5pU/Yzmepg31dTLouKFNCI918SJClwrWk5CSoCVBx9dqGyk6bRhc25HaEDiDM3iGgHuQNkOcEKLGqxXJEMC1rRvy975ohuvXou75BUWSISGqnqJUaqhadaIoCjqdjoKCgjKXr127ls6dO/PII48UD6U7cuRI8XIPDw/Cw8NZvnx58cQFZbnxxhsZOnQod955J3q9njvuuKPcddu0acPy5csZO3ZsqWX79+8nNTWVt956i7CwMAC2bt1aYh2TSbsuwWazlXqtPXr0oEePHrz88ss0bNiQ3377jQkTJpQbixC1VUpOEZP/jGXh7kRAu0nq8Pb1GNfRgxBSwHoCMk/BznlwbBWoZyZiCu8F7UaCV5iW+HiFaT1BMvJEiFql1iRDvZoF8DTdGM5arHt+xzjoLfnAEqIOKyoqIikpCYCMjAxmzJhBbm4uQ4cOLXP9Jk2a8O2337J06VIaN27Md999x5YtW0r0pkyePJmHH36YwMBABg8eTE5ODuvWrePxxx8vsa+bb76Z7777jlGjRmEwGBgxYkSZx3zllVfo168fjRs35o477sBqtbJ48WKeffZZGjRogMlk4uOPP+bhhx9mz549vPrqqyW2b9iwIYqisGDBAoYMGYKLiwuxsbEsX76cAQMGEBgYyKZNm0hJSSEyMvJyqlOIGsdutzN3UzzTlh0kq8ACwOCoIF7p60uwPRWsWZCbDLt+gMN/n0uCGnSDdneBT7g2C5xnffAKlRnhhKilak0y5O5kQG3Ul+y4GWeGym2GBl0dHZYQwkGWLFlCSEgIoPXqtGjRgvnz59OnT58y13/ooYfYsmULI0eORFEURo4cybhx41i8eHHxOvfccw+FhYV88MEHTJw4EX9//3ITnREjRmC32xk1ahQ6na7ElNln9enTh/nz5/Pqq6/y1ltv4enpSa9e2q0CAgICmD17Ni+88AIfffQRHTp0YNq0adx4443F29evX58pU6bwf//3f4wdO5bRo0fz3HPPsXr1aj788EOys7Np2LAh7733HoMHD77UqhSixtmfmM0Lv+3m3/hMAOp7uzB1SDj9fDPAfBLy02DXT9p9guxWbaP6HaH93eDXROsF9wjReoNkZjgharVakwwBXNsqjJij0dyiXwt7fpFkSIg6avbs2cyePfuC64SHh5eYbtrJyYlPPvmE7777rsSMc2+++WaJ7R566CEeeuihMvd5/vTVt912W/FMdOUZPnx4mYkSwMiRIxk5cuQFj/HSSy/x0ksvlShbsmTJBY8pRG1VaLHxfswBZq2Lw2JTMeoVRnUJY2InE66FJyE7E/bMhwOLwXbm2ruQttDubgg803vqHqjdPNVYsesEhRA1W61KhvpFBvF/f3TlFv1abHt+Rz/obRkqJ4QQQtQBKw4k88ofscSnaxMktAn14o3BYUSZTkNBBhyOgS1fgeXMDYgDW2o9QcFttOcuPtrQOJkdTog6pVYlQwEeTuTV70n26Zl45idDwiZo2M3RYQkhhBDiCknOLmTKX7Es2p2ECng4GRh/bQRjm1vQFcRDZjJsmAGn/tU28G0MHUZDvQ7acDgnDy0JcvF24KsQQjjKJXWbzJw5k4iICJydnYmOjmbNmjUXXH/u3Lm0bdsWV1dXQkJCGDt2LGlpaZcU8MVcGxVGjD1ae7LnlytyDCGEEEI4ls1m55sNcQz4YDULzyRC/VsGseyBFtwXnoYuP00bDvfno1oipDNC9Fi4/n2oH63NjBkYCfXaSSIkRB1W6WToxx9/ZPz48UyaNInt27fTs2dPBg8eTHx8fJnrr127ltGjR3PfffcRGxvL/Pnz2bJlC/fff/9lB1+W/i2DWWDTrhWyx/4OZ25SKIQQQojaIfZUFrf+bwOv/BFLZoGFet7OfHZHK77obyLEEg9ZJ2DZJNj4CVgKICASbvwYom4Bkyv4N4X6HcDtwjdGFkLUfpUeJvf+++9z3333FSczH374IUuXLuXTTz8tdaExwMaNGwkPD+eJJ54AICIigoceeoh33nnnMkMvW4S/G6f9u5Kd5Ypnfop29+iG3a/IsYQQQghx9RRabLy37ADfrD+O2WbHoFO4q0sYz3R1w73gJBRYYf9C+PcbsBaC3kkbEtfiBm1qbK/64N1QpskWQhSrVM+Q2Wxm27ZtDBgwoET5gAEDWL9+fZnbdO/enRMnTrBo0SJUVeX06dP8/PPPXH/99Zce9UVc2yqMZfaO2pPdP1+x4whRV5w/g5moXezSgy5qgJMZ+QyfuZ4v1hzDbLMTVd+Tn+9tw5SOFtzzT2q9QUtfgM3/0xKhoCitN6jlTdqkCCFtwLeRJEJCiBIq1TOUmpqKzWYjKCioRHlQUFDxzQ3P1717d+bOncvtt99OYWEhVquVG2+8kY8//rjc4xQVFVFUVFT8PDs7GwCLxYLFYrlonNc29+OjVV0YoV+Nfe+f2Pq/USs//M7WRUXqpK6RuinbpdZLSkoKfn5+KIpyJcKqFlRVxWw2U1BQUKtf53+pqorFYiElJQVFUVAUpdS5Ie+hcy63baorrsTn79bj6Tz+wy5Sc824GPU80SecMS0VDPlHseTZ0B1YgG7Htyg2M6rBGXv7MdibDQZFB65B2lTZOj048O8k7VL5pG7KJvVSvqqsk0uaTe78Lwqqqpb75WHv3r088cQTvPzyywwcOJDExESeeeYZHn74Yb766qsyt3nzzTeZMmVKqfIVK1bg6up60fhUFXYboshSXfHKT2H9z9NJc29RgVdWM8XExDg6hGpL6qZslakXk8mEr68vKSkpVzAi4SiqqpKfn09WVhYHDhwotTw/P98BUVVPl9s21TVV9fm7I01h7mEdZrtCgLPKAy2KCMo9wLLN4F6YSLv4L/HLOwRAintLdjS4j3xrAOzNOrOHDGBflcRSFaRdKp/UTdmkXkqryrZJUSsx/sVsNuPq6sr8+fO5+eabi8uffPJJduzYwapVq0ptM2rUKAoLC5k/f35x2dq1a+nZsyenTp0qvkP8f5X137ewsDASExPx8/OrUKxTFuyj7bZJ3GpYja39PdiHvFfRl1ljWCwWYmJi6N+/P0aj0dHhVCtSN2W71Hqx2WxYrdZaPVzOarWyfv16unfvjsFQq+46UC5FUdDr9ej1+nL/oZWWlkZISAhZWVl4enpe5Qirl6pom+qCqvr8tVrtvLHkAN9tSgCgfZgXn9wQRIAlEew2dPt+R7dzLordgmp0wdbhXtQmA7Xpst0CtOmy9dXnvSztUvmkbsom9VK+qmybKvUpYTKZiI6OJiYmpkQyFBMTw0033VTmNvn5+aW+WOj12pC18r5YOTk54eTkVKrcaDRW+GQYFFWPL7d05VZWozuwEP3QD2rlUDmoXL3UNVI3ZatsvdSFOrRYLFitVtzd3evE660oqYtzqqJtqksup14y8sw8+v121h/RbsMxokM93ujpjKkwCbLjYd2HkKb1BlGvA0q3xzC4B4LeBH5NwK36JqdyvpRP6qZsUi+lVWV9VPpfJhMmTGDUqFF07NiRbt268fnnnxMfH8/DDz8MwPPPP8/Jkyf59ttvARg6dCgPPPAAn376afEwufHjx9O5c2fq1atXZS/kfF0a+fK4qR2Zqhve+akQtx4a9bxixxNCCCHE5TuYlMMD323leFo+Bp3CswOa8GCzfChIgf0LYOtXYLeC0Q063QdN+p/pDfLXEiG9fGkUQlRcpZOh22+/nbS0NKZOnUpiYiJRUVEsWrSIhg0bApCYmFjinkNjxowhJyeHGTNm8PTTT+Pt7c21117L22+/XXWvogxGvY7eLeqxdE8nbjeshG1fSTIkhBBCVGMxe5OY8NNOcgqteLkY+XB4c/p6n4aCfNj8ORxYqK0Y2gm6PqolQDqDlgS5Bzg2eCFEjXRJg2nHjRvHuHHjylw2e/bsUmWPP/44jz/++KUc6rIMaBXMFzv7crthJer+BSh5qXKDNSGEEKKaUVWVmSuO8H7MQWyqSqMANz4f0ZgmJEB+Nqx+B05uAxSIHgOthmu9Qa6+WiJkKD18UQghKqJS9xmqaXo3CyBW34wd9kYoNgtsnOnokIQQQgjxHwVmK0/+sJ13lx3Apqr0bOrPb3c2oIk9DrJPweJntERI7wR9noeoW7SJEfybQlArSYSEEJelVidDbk4GejYJYLZ1kFaw7RuwyVztQgghRHVwKrOA2/63kT93JgIwtntDZl/viVfhSUjZDwufhsx4cPGFwW9Dw+7g7AX1OoBHsIOjF0LUBrU6GQK4sV09Ftq7koo35KfCrp8cHZIQQghR522NS+fmmevYfTILZ6OOt4ZF8kpHK/qCVDi2GpY8D4WZ4NsIrn9fGw7nEQLBrcHo7OjwhRC1RK1PhgZFBePu6sJ3ln5awabPHBuQEEIIUcf9tDWBUV9t5nR2EYEeTnw7Koo76qdCYTbsnKddI2S3QGhnGPS2dr2vbwT4N9GuFRJCiCpS65MhJ4Oe2zqF8b2tH1YMkLQL4jc7OiwhhBCizlFVlTcW7uO5n3dRYLERVc+T38a0oLNTAhTlwtr3YcdcbeWWN0HfSWByg8BI8Ap1bPBCiFqp1idDAHd1bkgq3vxp66oVrJ/u2ICEEEKIOmjyn7F8vuYoKnB96xB+uiOU+uZjkJ8OyybB0RWg6LRpszs9oA2HC24tM8EKIa6YOpEMNfBzpWezAGZbB2oFB5dAdqJjgxJCCCHqkPeXHeCbDccBeLRPI2b0d8E17wRkJWgTJSTv1W6ket0UaD4YjK4Q0hacPR0cuRCiNqsTyRDAqK4N2aU2ZidNtTtXb/jE0SEJIYQQdcLXa4/y0T+HAXjwmoY806ZIu/df4g5YNBFyk8A9CIa8C/XaazPGhbQBo4tjAxdC1Hp1Jhm6tkUgIV7OfGk+0zu0Yy5YzY4NSgghhKjlft6WwKsL9wFwW3Q9nm9XCEU52iiNmFfAnAcBkdqMcd4NwD0QgqJAb3Rw5EKIuqDOJEN6ncJdXRqw2N6ZNMUHCtLPXaQphBBCiCq3LDaJ//tlN6oKg1oF8lZ3FaUoF7Z+DRtmgGqDiN4w8HWtN8g7DAKag67OfD0RQjhYnfq0ua1TGKrOyCzzdVrB5v85NiAhhBCiltp0LI0nftiO1a7SvbEvH/dzQpefDivfhNhftZXa3gk9J4LBCfybgU+4Q2MWQtQ9dSoZCvRwZlCrYObZrtWm2U7eB0dXOzosIYQQolZJyodH5u6k0GKnbagXXw3xxFiQDqvfhYSNoDNoSVC7O7XhcEGtwCPI0WELIeqgOpUMAdzdtSFpePGXvbtWsGGGYwMSQgghapFjKbnM3Kcnp8hK00B3vhkWgEtRmtbeJmwEnRGumwqN+oDBWZsowcXH0WELIeqoOpcMdW3kS+MAN760nJlI4fDfkBHv2KCEEEKIWuBUZj73fLONLLNCqI8Lc26tj7clGf79Bg7HaPcQ6v2slgA5uWtTZ5vcHB22EKIOq3PJkKIo3N21IbFqBLt1LbSLN6V3SAghhLgs6XlF3PXlZhKzivB1UvlmRChB1kTt+qA9P2srdXsMGnQDVz8IbgMGk2ODFkLUeXUuGQIY3iEUZ6OOTwsHaAW7fgRzvmODEkIIIWqovCIro77azLHUPLxcDDwSaaOBmgRHlmszxwF0uAeaDgCPYAiMBJ3esUELIQR1NBnycjFyY9t6LLN3JEPvB4WZWhe+EEIIISqlyGJjzKzNxJ7KxtWk538jGhHoAsqJzbBuurZSy2EQNULrEfJrAori0JiFEOKsOpkMAYzqGo4VA18VnZlme+vXoKqODUoIIYSoQWw2Ow/P2caWuAxMeh2f3NqcaOckfHMPoF/zNqh2aNQXOt4LLt4Q0EISISFEtVJnk6HWoV60ru/FXGtfrIoJUg9qkykIIYQQ4qLsdjsT5u9kxYEU9IrCu8Oa0tczCdKP0eXoByg2M4R2gh5PgpMHBLaUm6kKIaqdOv2pNKpbQzLwZLFyjVawcaZjAxJCCCFqiCl/7eWPHadQgFeub8JNQWmQdQLDP69gsuVjD4iE3s+ByR2CokBvcHTIQghRSp1Ohoa2qYeHs4GZBf21gqMrIeWgQ2MSQgghqrv3lx3gmw3HAXj6usaMbpgFuckQ8xJKQTrZzqHY+rys9QgFR8mscUKIaqtOJ0MuJj23RoeyT23IfmMrbWzzxk8cHZYQQghRbX27IY6P/jkMwH09GvBYZB7kp8Pfr0BOIqpbIBsaTwQXTwhqBUYXB0cshBDlq9PJEMBdXRsC8GHemd6hPb9CQabjAhJCCCGqqV0nMpn6114AbmlfjxejbVqbueI1SD8Czl5Y+02l0OQLAc21niEhhKjG6nwy1DjAna6NfImxR5NlDICibNg2y9FhCSGEENVKocXGE/O2Y7WrdI3w5Z1eepSCTFg9DZJ2gcEFrpsCnvW1DZy9HRmuEEJUSJ1PhgBGdwvHhp7ZljO9Q/9+BzarY4MSQgghqpFX/oglLi0fbxcjHw/wRF+QqU08FL8edAa49kXtHkK+jRwdqhBCVJgkQ0D/lkEEuJv4urAPNp1J6+o/sNDRYQkhhBDVQszeJH7cmgDAm4PrEUA6bJ8Dh5YCCvR6BkLagncD8AhybLBCCFEJkgwBRr2OkV0akoU7/xh7a4Wbv3BsUEIIIUQ1kJpTxLM/7wJgRLsgBgdmwd4/YPeP2gpdx0HDHuARAj4NHRipEEJUniRDZ4zsHKbdNC77Oq3g+DpI2u3YoIQQQggHUlWV8T/uICPfQpiPC692U+DIcthy5h+G7e6G5oPBzR/8Gjs2WCGEuASSDJ0R4uVC3xaBHFTDOOrS+sw02586OiwhhBDCYWavj2Pt4VQMOoWPrw/EJW0vrPtIWxg5FNrcDi7e4N8cFMWhsQohxKWQZOg/RnfTuvffzxuoFez7E3JTHBiREEII4RiHT+fw1uL9ADzWsz7tnE/Dmmlgt0D9aOj0ADh7QEAk6OTrhBCiZpJPr/+4pok/YT4uLDK3I9cpCIpyYOvXjg5LCCGEuKosNjuPfr+dIquddqGePNGqELZ9Axlx4OwFPcaDyQ0CW4He4OhwhRDikl1SMjRz5kwiIiJwdnYmOjqaNWvWXHD9oqIiJk2aRMOGDXFycqJx48Z8/XX1SzJ0OoVR3RpiR8c8BmmFO+aCpdCxgQkhhBBX0ZuL9nPgdA5uTno+HuiFLn4D7PtDW9hjPLgHQVArMJgcGqcQQlyuSidDP/74I+PHj2fSpEls376dnj17MnjwYOLj48vd5rbbbmP58uV89dVXHDhwgHnz5tGiRYvLCvxKuTU6DJNex8dZ12DTO0Hmcdjzi6PDEkIIIa6KdYdTmbX+GABT+ocSZo2HtR9oC1vcAGFdtETI6OLAKIUQompUOhl6//33ue+++7j//vuJjIzkww8/JCwsjE8/LXuygSVLlrBq1SoWLVrEddddR3h4OJ07d6Z79+6XHfyV4ONmYkjrYLJxY6NrX63w32+1IXNCCCFELZZVYGbCTztQVRjU0p8RoVmwbjoUZoJXA4geq91U1cnd0aEKIUSVqFQyZDab2bZtGwMGDChRPmDAANavX1/mNn/++ScdO3bknXfeoX79+jRr1oyJEydSUFBw6VFfYaO6hQPweno/rSBhExxZ4biAhBBCiKvg2Z93cTq7iCBPJ97paYD9C+DEZtAZoNdE8KwHniGODlMIIapMpa56TE1NxWazERRU8u7SQUFBJCUllbnN0aNHWbt2Lc7Ozvz222+kpqYybtw40tPTy71uqKioiKKiouLn2dnZAFgsFiwWS2VCviStQ9xoFuTO3tP1ifPvTHjWZuzrZ2Cr3xlc/a748SvqbF1cjTqpaaRuyib1Uj6pm7JJfZzj6LbpSvt1+0mWxp5GUWDawEBc0nejbvkSBbC1vwe7X3PwCoeLvFZ5L5VN6qV8Ujdlk3opX1XWySVNAaOcdy8BVVVLlZ1lt9tRFIW5c+fi5eUFaEPtRowYwSeffIKLS+kxx2+++SZTpkwpVb5ixQpcXV0vJeRKa+emcBA9z2aN4Ae2ojuxifULviHNvfpd6xQTE+PoEKotqZuySb2UT+qmpPz8fEeHUG1Uh7bpSskogrd36gGFAfXsZMYfJO/gW3jbzCR7RLHB1hN2noadFX9/yHupbFIv5ZO6KZvUS2lV2TZVKhny9/dHr9eX6gVKTk4u1Vt0VkhICPXr1y9OhAAiIyNRVZUTJ07QtGnTUts8//zzTJgwofh5dnY2YWFh9O3bFz+/q9Mz07vIyoJ3VrG5MJTExgOof3IJ3dPnYxvwE3iHXZUYLsZisRATE0P//v0xGo2ODqdakbopm9RL+aRuypaWluboEKqN6tA2XQk2u8rtX2yiwJZN8yA3PhjijNPWz9AXxKM6eeDT/xmGBEeBT8MK7U/eS2WTeimf1E3ZpF7KV5VtU6WSIZPJRHR0NDExMdx8883F5TExMdx0001lbtOjRw/mz59Pbm4u7u7aBZcHDx5Ep9MRGhpa5jZOTk44OTmVKjcajVftZPA2Grm5fX3mbornA/tIphlXoUs7hG7PD9BzYrWaTvRq1ktNI3VTNqmX8kndlCR1cU51aJuuhI9jDrLzRDbOBh2fDPHDNWkl7PsNAKX7kxj9G4J/o0rfWLWm18uVIvVSPqmbskm9lFaV9VHp2eQmTJjAl19+yddff82+fft46qmniI+P5+GHHwa0/5yNHj26eP0777wTPz8/xo4dy969e1m9ejXPPPMM9957b5lD5KqT0WcmUvg1zkhmsxFa4dZZkHrIcUEJIYQQVWR7fAYzVxwG4P/6hdLEHg9r39cWNhsE4T3Av3mlEyEhhKgpKv3pdvvtt/Phhx8ydepU2rVrx+rVq1m0aBENG2rd54mJiSXuOeTu7k5MTAyZmZl07NiRu+66i6FDh/LRRx9V3au4QpoHe9C1kR92FV7NvUmbPCEvGbZ+CeY8R4cnhBBCXLICs40n5m3Halfp2diXe8KzYcPHUJAOnqHQ6X7wbQymmn09lBBCXMglTaAwbtw4xo0bV+ay2bNnlypr0aJFjb346/nBLbjpk3X8esDM011HU2/HB7DzR4gaof3HTAghhKiBXvx9NwkZBfi6GvngWieUg79C/IZz02h7hYJH2dcDCyFEbSH93hfRNsybga2CUIH/S+oFPhFgyYNNn0F+uqPDE0IIISpt8e5Efvn3JADvDKmHf84B2PK5trD9KAhqBX6lJzgSQojaRpKhCvi/wZHodQqr4wrY2/h+rXD/Qji2FlTVscEJIYQQlZCaW8Tzv+4G4I4OwVznkwJrpoG1CILbQNRw8G8G+ksaPCKEEDWKJEMVEOHvxm0dtZnvnjncCrV+NKg22PwZ5CQ6ODohhBCi4ib/GUtmgYVwP1cmd1Zh+xxIOwwmd7hmAng3ABdvR4cphBBXhSRDFfRU/2Y4G3XEJuWzOuR+UHRwfB0cWAw2q6PDE0IIIS5q76ksFu7S/ok3baA/zombYM8v2sLuj4NvBHhX7H5CQghRG0gyVEGBHs7c2yMCgJdiA7E3vk5bsPlzyDzuwMiEEEKIinl94T5UoH8zbzqaEmDN+4AKTQdARC8IaA6K4ugwhRDiqpFkqBIe6dMYLxcj8RlF/OZ9DxicIGU/7P4ZLAWODk8IIYQo1/rDqaw7koZep/BCRwU2zID8VPCsB50eAL8mYKze9/8TQoiqJslQJXg4G3n82iYAvLHdhDnyFm3B1q/lRqxCCCGqLVVVeW3hPgBGtPYh4vRSbai3ooeeE8GnIbgHOjhKIYS4+iQZqqRR3RoS4uVMWp6FL3UjwMUHcpPg32+hINPR4QkhhBClLNyVyN7EbJyNOp5tchI2/U9b0P4ubQY538aODVAIIRxEkqFKcjLoeXZgcwBmbLeQ1+pubcGuHyBplwMjE0IIIUqz2Oy8vWQ/APdG++C3YyZYCyAoSruBeEBzmUZbCFFnSTJ0CW5qV59mQe7km228mzdQm3mnKAe2zoKc044OTwghhCg2b1M8CRkFeLsYeMxnC5zcpg2POzt7nLOno0MUQgiHkWToEuh0CpOGRAIwd2cOKa3u1Rbs+xMSNoLd5sDohBBCCE2+2cr05do1rY939sB1x5faghbXQ2AkeIU5MDohhHA8SYYuUa9mAXQO98ViV5lysiOEtAO7FbZ8BVknHB2eEEIIweerj5KWZybE04l79EshIw6MbtDubvCXabSFEEKSoUukKAov3qD1Di3cn8XR5g8AChxbBUf+AWuRYwMUQghRp6Xnmfl89VEA/q+zAcP2b7UFbW6Deu3A6Oy44IQQopqQZOgytAn1ZkDLIFTg5QMNoHE/bcHmLyA9zpGhCSGEqOM+/Psg+WYbTQNcGJo7HwrSwS0Q2t4h02gLIcQZkgxdpueHRKLXKaw9lsO2sHtA7wTJsbDvD21SBSGEEOIqS0jP5/vN8QBMiS5Et+cXbUGH0dq1QjI8TgghAEmGLluEvxu3RocC8PIOD9SWN2kLts2GlIOOC0wIIUSd9faS/VhtKp3D3Ol26lttKm2/phA1HJy9HB2eEEJUG5IMVYEJA5rhbNQRm5TPUp87wdkbsk/C7p8gL9XR4QkhhKhD9p7KZuGuRACmtkpEObRUW9DpfvBt5MDIhBCi+pFkqAoEejhzb48IAF7fYsfWZqS2YMf3kLQbbFYHRieEEKIueX3hXlRgYDMvWhz+ClQ7hHWFZgPA6OLo8IQQolqRZKiKPNKnMd4uRhIyCvleNxS8QqEoG3bMhbTDjg5PCCFEHbDhSCrrjqSh1ylMabgLTmwGRQed7wevBo4OTwghqh1JhqqIh7ORx65tAsAHm/MoanfmRqx7/4DTsZCb7MDohBBC1HaqqvLawn0A3B7lSXDsmRusNhsMDXuA3uDA6IQQonqSZKgKje4WToiXM+n5Fj7N6QFBUWAzw7ZZWu+QpcDRIQohhKilFu1OJPZUNs5GHf/nswLSj2jD4jreCx4hjg5PCCGqJUmGqpDJoOPZgc0B+HxLJlltH9CGJ8StgcP/QMoBUFUHRymEEKK2sdjsvLXkAAAPtXPGc8832oLWt0L9DjKVthBClEOSoSp2U7v6NAtyJ99i470TzaHN7dqCjTMh7QhkHndsgEIIIWqdHzbHk5Cej7eLgXH63yEvBVz9ocMYcPV1dHhCCFFtSTJUxXQ6hUlDIgH4fkcGCU3uhoDmYMmDNdMg4zgUZDo2SCGEELVGvtnKB38fAmBieztOsT9pCzqMhsAWDoxMCCGqP0mGroBezQLoFO6D1a7y9r866PmMNm47eS/sng+pB2W6bSGEEFXi89VHSc8zU8/TxMi8OWDJ1+4n1HYkmNwcHZ4QQlRrkgxdAYqi8OL1Wu/Qwn0ZxNIYujyiLdwxFxJ3QdohB0YohBCiNkjPM/P56qMATGmbgf7gYm1BpwfBN8KBkQkhRM0gydAV0jbMhwEtg1CBF9fkY28yAMJ7aTe/WzMNMuMhJ8nRYQohhKjBpv99kHyzjeaBLlx3+mtQbVC/I7QcCnqjo8MTQohqT5KhK2jS9ZE4GXRsT8hi5gFX6DYO3AIgJxE2fw7pR8Gc7+gwhRBC1EAJ6fl8vzkegLebH0aJ36DNYNrlYfCo5+DohBCiZpBk6Apq6OfGM2em2v5w1Ul2W8Pgmqe1xurw33B0FaTKdNtCCCEq750l+7HYVLqEudHu2BdaYdMB0Lgv6KR5F0KIipBPyyvsvmsi6N7YD6td5fFFKRTW6wRRt2oLN3wM6ccgI86hMQohhKhZ9p7KZsGuRADeDVuvTcxjcIYuD4Gbv4OjE0KImkOSoStMURQ+vL0dXi5G4tLymbxJgfZ3gV9TMOfB2vfPTLed4ehQhRBC1BBvLNqHCtzQ1IUGB2drhVG3QGgnR4YlhBA1ziUlQzNnziQiIgJnZ2eio6NZs2ZNhbZbt24dBoOBdu3aXcpha6xAT2feuqU1AD/8m8Tf6X7Q6xntv3hJuyH2N0g9BDaLgyMVQghR3e1IyGTt4VT0OoWp3gsg9zS4+ELnB8HJw9HhCSFEjVLpZOjHH39k/PjxTJo0ie3bt9OzZ08GDx5MfHz8BbfLyspi9OjR9OvX75KDrckGR4UwokMoAM8sOkWaRzPo/JC2cPt3cDpWS4iEEEKIC/hmfRwAtzVR8T3wg1bYYTQERjouKCGEqKEqnQy9//773Hfffdx///1ERkby4YcfEhYWxqeffnrB7R566CHuvPNOunXrdsnB1nSv3RxFmK8LGfkWxi8vRG02CBr20KZCXf0uZJ2E7ERHhymEEKKaysgzs3C31k6M1/2gDbf2CYcO94DBybHBCSFEDVSpZMhsNrNt2zYGDBhQonzAgAGsX7++3O1mzZrFkSNHeOWVVy4tylrC2ajn45EdMOgU1hxJZ/YxD+j2GLj6QfZJ2PKFTLcthBCiXPO3JWC22unjl0Fg/CKtsNOD4NPQsYEJIUQNZajMyqmpqdhsNoKCgkqUBwUFkZRU9g1EDx06xP/93/+xZs0aDIaKHa6oqIiioqLi59nZ2QBYLBYslpp9XU2rYDce6R3BxyuO8tbyE3QdWY+m3Seg//tFlENLsYZEo+qcIKj1RadGPVsXNb1OrgSpm7JJvZRP6qZsUh/nOLptsttVvttwHIDJpu9R8qzYQ9pjixwGNrv2qAbkvVQ2qZfySd2UTeqlfFVZJ5VKhs5SFKXEc1VVS5UB2Gw27rzzTqZMmUKzZs0qvP8333yTKVOmlCpfsWIFrq6ulQ+4mmmkQlNPHYey4aFfTzA+KozWgUNomrwQ+7rprEgPotB4ssL7i4mJuYLR1mxSN2WTeimf1E1J+fnSU32Wo9um/ZkKCRl62ujjCM9Yh4rCKq9byF699Yof+1LIe6lsUi/lk7opm9RLaVXZNimqWvE7fprNZlxdXZk/fz4333xzcfmTTz7Jjh07WLVqVYn1MzMz8fHxQa/XF5fZ7XZUVUWv17Ns2TKuvfbaUscp679vYWFhJCYm4ufnV6kXWF2dyCjghhnryTPbGNs5mBeiMjEseQYl/Qj24HbY+k2BoJbg4lPuPiwWCzExMfTv3x+j0XgVo6/+pG7KJvVSPqmbsqWlpRESEkJWVhaenp6ODsehHN02PTJ3O3/vT+HngK/omLMce8Oe2EbMBmevK37sypD3UtmkXsondVM2qZfyVWXbVKmeIZPJRHR0NDExMSWSoZiYGG666aZS63t6erJ79+4SZTNnzuSff/7h559/JiIioszjODk54eRU+kJQo9FYa06GiEAjU26KYuL8nczeksR1EeH06PUM/PUkuqQd6A78AU4u4OoNBtMF91Wb6qWqSd2UTeqlfFI3JUldnOPItikpq5AVB1LxJZsOeasB0HUYhc6j+t5gVd5LZZN6KZ/UTdmkXkqryvqo9DC5CRMmMGrUKDp27Ei3bt34/PPPiY+P5+GHHwbg+eef5+TJk3z77bfodDqioqJKbB8YGIizs3Op8rpoRHQof+87zZI9SUxYlMiyO5vj1fkB2DAD/v0WQtqB0blC1w8JIYSoveZtPo5NVZnosxpdgQX8m0HkUEeHJYQQNV6lv2HffvvtfPjhh0ydOpV27dqxevVqFi1aRMOG2kw2iYmJF73nkDjnnVtaE+TpxOmcIp5dY4Hm10NYV7Bbtem2c5MhZR9UfDSjEEKIWsRis/P95gQMWLnZtlQrbH83mGr+NbRCCOFol9TdMG7cOOLi4igqKmLbtm306tWreNns2bNZuXJludtOnjyZHTt2XMphayVPFxMf3NYORYGle1OZf8IDuj+u3U08KwG2zoL8dEg77OhQhRBCOMDyfadJySniFud/cTGngbM3tB/l6LCEEKJWkLFX1UD3Jv7c20O7fmpyzEkSjA3hmqe0hQcWwrFVkJME6cccGKUQQghH+PbMdNqPuyzRCloNB1dfB0YkhBC1hyRD1cT/DW5Bi2AP8opsPL40C3tYF63BA1j7AZzaDlknIKviU24LIYSo2Y6l5rH+SBqtlaOEFuwHnQG6POzosIQQotaQZKiaMOp1zBjZHmeDjh0JWXwU6wLRYyC8p3b90IrXIfUgpB/VriMSQghR683ZqPUKPef1t1bQqA8ENndcQEIIUctIMlSNNAny4LnBLQCYseYE2/P94ZoJENIerIXw92TtOqLUg9p1REIIIWqtQouN+VsT8CeL7kVrtcIuDzk2KCGEqGUkGapmxnQPp2cTf6x2lScWnqbA6AV9nwe/plCUDTEvQ24KpOyHohxHhyuEEOIKWbArkexCKw+6rkCnWiEgEhpf5+iwhBCiVpFkqJpRFIUPbm+Hr6uRhIwCJm0AXP3gusngGQp5KRDzEuRnaAmREEKIWunbDXEYsXKXLkYriL5H7jknhBBVTD5VqyF/DyfeGdEWgF93JDH3uBe4BUD/qVpilJUA/0yBolxtA6vZgdEKIYSoantOZrHrRBY3GDbjZs3Qbrcg02kLIUSVk2SomrquZRAjOzcA4OWFh1iY6g+e9aD/q+DkASkH0K96E8VuheS9YLM4OGIhhBBV5ezECU+4nrnJapvbwMndgREJIUTtJMlQNfbasCh6NvXHZleZ8OtB1uSEgE849HsFDE7oEv+lQ/znYM6D07Fgtzk6ZCGEEJcpu9DC7ztO0k45TIT5kDaddtdxjg5LCCFqJUmGqjG9TuF/o6Jp38CbIqudh+fvZ7u5PgRGQp8XUBU9oRkb0W39AgqztWuIVNXRYQshhLgMv247QaHFzmOuZ64VatIffBo6NighhKilJBmq5lxNBr4Z24nmQdoNWe/94SCH7KFQPxpb9/EA6A8sgN0/adNtpx50bMBCCCEumaqqfLvxOAFk0Ne+QSvs+ohjgxJCiFpMkqEawNPFxHf3daaBrysZ+Rbu/uEw8bp6qBF92F3/bm2l7d/BgcXaDVnTjzo0XiGEEJdm07F0jqbkMca0HL1qhaAoaNTb0WEJIUStJclQDRHo6cyc+zsT6OHE6ewiRv0UT6ohiKOBA7BF3a6ttHEmxK2FrJOQmeDYgIUQQlTadxuOY8LCaONyraDTfY4NSAghajlJhmqQBr5ufHdfF7xdjRxPy2fs70kU2cDe9i5oNghQYc00SNwJGXGQc9rRIQshhKiglJwilsYmcb1uIx62LHDzh7Z3OTosIYSo1SQZqmGaB3vw9ZhOuJn07E/K5esDOgqdfKHLI9CwO9it8M9rkHYY0g5JQiSEEDXET1sTsNrtPOpyZjrttneC0cmxQQkhRC0nyVAN1KGBDzPvjsbJoGN/lo5HY/KwuvhCz2cguA1YC+DvVyDzhDahQkaco0MWQghxATa7ypyNx+mgHKKJ7SjojDKdthBCXAWSDNVQvZsF8O4tUegVldWH0hi/woLq4gPXvgh+TaAwC2JegrxU7fqhlANgtzs6bCGEEGVYeSCZxKxCHjSd6RVqPhg8QxwblBBC1AGSDNVgg6OCGdXEjgIs2J3ES5t12hjzfpPBsx7kJWsJUW6y9ji9B2wWR4cthBDiPN9tOE4waQxQNmkF3R9zbEBCCFFHSDJUw7X3V5k4oCkAczadYNouZ/AIhv6vgqsfZCXAwglaz1BhFiTuAkuBg6MWQghxVkJ6PqsOpnCXYTk67BDSDsK6ODosIYSoEyQZqgUe7BnBQ70bATBjVRxfHvEArzAYMg18IqAwE5Y+r027bcnXZpsrzHZs0EIIIQD4fnM8JszcY/xHK+jysGMDEkKIOkSSoVri/wa14I5OYQC8vvQI80/5gkcIDH4bQjuBzQyr3oLdP4HVDEm7teuJhBBCOEyR1caPm+O5Ub8eTzUb3AOh9a2ODksIIeoMSYZqCUVReHN4awZHBaOq8Pyfh1iWHgROHtD3RYi8SVvx329h3YdgLYLkfXJzViGEcKAle5JIzzdzv3GZVtB+NOgNjg1KCCHqEEmGahFFUfh4ZHuuaeKP1a7y+C8HWJNXH0xu0PkB6DIOFB0cWa5NrFCYrU27nXoYVNXR4QshRJ3z3cbjdFIO0Jw40DvJdNpCCHGVSTJUyxj0Or4YHU27MG+KrHYe+mEfvycHaD1ELYZAv1fA6KrNLLfoacg6CTmJcDoW7DZHhy+EEHXGgaQctsZlMNawRCuIvAHc/BwblBBC1DGSDNVCLiYD397bmZYhnuSbbTz1cyzv7HLC7uoP9aNh8LvgHqQlQYuehqRdUJChTaxgLXJ0+EIIUSfM3XSceqQyUL9VK+gx3qHxCCFEXSTJUC3l6WJk/sNduS4yEFWFmSuP8tDSPHLdQsGnIQx5DwKagzkXlr0Eh5aBOU9LiIpyHR2+EELUanlFVn7ZdoJRhhj02LWJbkLaODosIYSocyQZqsXcnIz87+5oxvVpjKJAzN7T3PJ9AvGmCG0oxoA3ILwXqDZY/xFsm63dgyhpF+SnOzp8IYSotX7fcRKbOZ87DSu0gq6PODYgIYSooyQZquX0eh3PDmrBh7e1w82k50BSDjd/vY8NBWHg7AW9noG2I7WV9/wMK9/SeoiS90L2KccGL4QQtdQPmxO4Sb8eL3K12yC0HObokIQQok6SZKiOuKl9fb5/sCv1vJ1JyzNzz3d7mJPgCy5e0O4uuOZp0Bkgfj0seR7y0iDtiDaxgtXs6PCFEKLWOJlZwO6TmYzVn5k4oeO9oNM7NighhKijJBmqQ9qGevP7uB50aOCN2WbnxT/38+JGBatrADTuCwNeBydPSDsECydA+lFtuNypfyE3xdHhCyFErRATm0RX3T5a6BLA4Kzd+kAIIYRDSDJUxwR6OvPDg90YER0KwJxNCdz9RwYZzvUhOEqbWMErFPJTYfGzELcGbBZI2Q/J+7XfhRBCXLJle09zj36p9qTVcHDxcWxAQghRh11SMjRz5kwiIiJwdnYmOjqaNWvWlLvur7/+Sv/+/QkICMDT05Nu3bqxdOnSSw5YXD6TQce0W9syaUgLDDqFjUfTGfbdcQ6oDcA7FAZPg5C2YC2EVW/DyjehIBPyUuDkvzK5ghBCXKLMfDN7jybQT/evViA3WRVCCIeqdDL0448/Mn78eCZNmsT27dvp2bMngwcPJj4+vsz1V69eTf/+/Vm0aBHbtm2jb9++DB06lO3bt1928OLyPNCrMV/d0xFvVyPH0/MZ8c0+lmXWAzd/uG4KtLkdFB0cXwd/jIOjq7T7EJ2OhZSDYLM6+iUIIUSN8s/+ZPopWzEpNvBrAiGtHR2SEELUaZVOht5//33uu+8+7r//fiIjI/nwww8JCwvj008/LXP9Dz/8kGeffZZOnTrRtGlT3njjDZo2bcpff/112cGLy9e7eSC/PdKDxgFu5BRaeeSHvcw44A6uvtB+FFz/AfhEQFE2rHkXVryu9QzlnoZT27WbtQohhKiQZbGnuV6/UXvS6hbHBiOEEAJDZVY2m81s27aN//u//ytRPmDAANavX1+hfdjtdnJycvD19a3MocUVFBHgxh+P9uDR77ez6mAK0/4+yoHkYKb1DcTJD7j+fdg9H3b9CAkb4fQe6PwgNOoLSXvAsx74hMtsSEIIcQGFFhvbDx7jI91uraC1JENCXA6bzYbFUnuvZbZYLBgMBgoLC7HZbI4O56oyGo3o9Vfne2WlkqHU1FRsNhtBQUElyoOCgkhKSqrQPt577z3y8vK47bbbyl2nqKiIoqKi4ufZ2dmAdlLU5pO+ss7WRVXUiZMePr+rLe8uO8RX647z164kjqV6MmNYA+oridB6JIR2xbBhOkr6EVj7PvZjq7F1eQxsKuSmaUM+nDwuO5aqUJV1U5tIvZRP6qZsUh/nXG7btHJ/Mj3tmzHpbdj9mmLzbgS1sH7lvVQ2qZfyVbZuVFUlOTm5+D1YW6mqSnBwMPHx8SiK4uhwrjpPT08CAwPLfO1V+T6qVDJ01vlBqapaoT/SvHnzmDx5Mn/88QeBgYHlrvfmm28yZcqUUuUrVqzA1dW18gHXcjExMVW2r9bAfc0VvjukY8+pbK7/fDc3h9vp6K+iKN4oDSbRxGkRzZN+R39yK7bfH2FP6J3E+/YCJbHK4qgqVVk3tYnUS/mkbkrKz893dAjVxuW2Td8f1vGwThsid8DUmoOLFlV5jNWJvJfKJvVSvorWjYeHBz4+Pvj7+2MymepkolCbqaqK2WwmJSWFgwcPkpOTU2qdqmybFFVV1YqubDabcXV1Zf78+dx8883F5U8++SQ7duxg1apV5W77448/MnbsWObPn8/1119/weOU9d+3sLAwEhMT8fPzq2i4tZ7FYiEmJob+/ftjNBqrdN87T2Ty1E+7ScgoAKB1PU+mDm5IlNNpMBdA5nH0Gz5Cl3YQAHtIe2xdHwO3QDC5gF9TMLlVaUyVcSXrpiaTeimf1E3Z0tLSCAkJISsrC09PT0eH41CX0zbZ7CoD3lrAcvv9GBUbloc2gH/TKx2yQ8h7qWxSL+WrTN3YbDaOHj1KQEBArf9OqKoqOTk5eHh41MmELy0tjZSUFBo1alRqyFxVtk2V6hkymUxER0cTExNTIhmKiYnhpptuKne7efPmce+99zJv3ryLJkIATk5OODk5lSo3Go3yAVKGK1EvHSMCWDK+F+/HHOTbDXHsPpXNbbP3cFfnBkzs6oe7QQ9D3oW9v8P2OegSt6Nb8BhE3wvNBkHKHvCsD15hoL+kDsgqIedM2aReyid1U5LUxTmX0zZtP5ZOF/MGjEYbqn9zjCEtr1SY1Ya8l8om9VK+itSNzWZDURTc3d3R6Wr37TLtdjugjciq7a+1LO7u7qSmpgKl26KqfA9VumYnTJjAl19+yddff82+fft46qmniI+P5+GHHwbg+eefZ/To0cXrz5s3j9GjR/Pee+/RtWtXkpKSSEpKIisrq8pehLgy3JwMvHRDS34b14N2Yd5YbCqzNxxn4Kw4lmXXB1cfiLoFbvwYAiLBUgAbP4FlkyA7EbJOwMmt2s8zb2ghhKiLlsUmcb1uEwBK1AgHRyNEzVcXe0rqmqv1N650MnT77bfz4YcfMnXqVNq1a8fq1atZtGgRDRs2BCAxMbHEPYf+97//YbVaefTRRwkJCSl+PPnkk1X3KsQVFVXfi18f6cbkG1vi5WLkZGYBD34fy0N/W0gyhWkzyQ16Czo9AHonSNoFfz4G+/4CSyGkH9OSopzTUPFRmUIIUSuoqsqG2MP00O3RCqJuvvAGQgghrppL6nMbN24ccXFxFBUVsW3bNnr16lW8bPbs2axcubL4+cqVK1FVtdRj9uzZlxu7uIp0Oh1jukew7KleDI4KBmBp7GkGfHGAWcf9sLv5Q8ubtF6ioCiwFsLm/527WaulAFIPwql/IS/Nwa9GCCGunv1JObTMXoNRsWH3b1FrrxUSQlxZs2fPxtvb29Fh1Dp1bwCiuCxBns58enc0X4/pRKiPC9mFVqYsOsTNP2ey136ml2jgG9DlEXDyhOyT2s1a/3wMjq+HojxI3guJO6FQhkoKIWq/ZbGnueHMLHK61jJEToi6LCEhgfvuu4969ephMplo2LAhTz75JGlpJf9RHB4ezvTp0x0UZdXo06cP48ePd3QYFyXJkLgk17YIJOap3jzQMwKjXmHniSxu+novr+10o9AlGFpcD7d8Ce1Ha7PKZcbDyjdgwXhI2AwFWZC4C07HgjnP0S9HCCGumA17Dp4bItdKhsgJUVcdPXqUjh07cvDgQebNm8fhw4f57LPPWL58Od26dSM9Pd0hcdX1e19JMiQumYtJz6TrW/LnY9fQ/swEC1+uO851355kRXZ9cPOHNrfBLV9B25FgdIH0I/DPVFg0EU5t14bMndoOKQe164uEEKIWOZGRT4OUFRgUO1b/SPBv4uiQhBAO8uijj2IymVi2bBm9e/emQYMGDB48mL///puTJ08yadIkQOtROX78OBMmTMDHx6fUtNJLly4lMjISd3d3Bg0aRGJiyfs8zpo1i8jISJydnWnRogUzZ84sXhYXF4eiKPz000/06dMHZ2dn5syZU2a8mZmZPPjggwQFBeHs7ExUVBQLFiwAtKmtR44cSWhoKK6urrRu3Zp58+YVbztmzBhWrVrF9OnTURQFRVGIi4sjIyODu+66i4CAAFxcXGjatCmzZs2qkvq9VI6b81jUGpEhnvw6rjvfbTzOe8sOciKjgLHz9jGoVRATezagiXsqtLsLWgyF2F9h/1+QegBiXtKuL2p3NwRHQV4KeASDdwPQy7SjQoiaL2bvuSFyhjYyRE6IK0FVVQosNocc28Wor9CsZ+np6SxdupTXX38dFxeXEsuCg4O56667+PHHH5k5cya//vorbdu25YEHHuD222/Hw8OjeN38/HymTZvGd999h06n4+6772bixInMnTsXgC+++IJXXnmFGTNm0L59e7Zv384DDzyAm5sb99xzT/F+nnvuOd577z1mzZpV5i0D7HY7gwcPJicnhzlz5tC4cWP27t1bnJgVFhYSHR3Nc889h6enJwsXLmTUqFE0atSILl26MH36dA4ePEhUVBRTp04FICAggCeffJK9e/eyePFi/P39OXz4MAUFBZWv+CokyZCoEoqiMLpbOIOjgnnlz70s2p3IktjTxOxLpn9kEI93D6OVZzpEj9EmWtj9MxxYBKf3wNL/g5B20P5uUO2Qmwxe9cGjnkPvUSSEEJdrw66DjNLFak9kiJwQV0SBxUbLl5c65Nh7pw7E1XTx7yqHDh1CVVUiIyPLXB4ZGUlGRgYpKSkEBgai1+vx8PAgKCioxE1FLRYLn332GY0bNwbgscceK042AF599VXee+89hg8fDkBERAR79+7lf//7X4lkaPz48cXrlOXvv/9m8+bN7Nu3j2bNmgHQqFGj4uX169dn4sSJxc8ff/xxlixZwvz58+nSpQteXl6YTCZcXV0JDg4uXi8+Pp727dvTsWNHQLs2ytHkm6aoUgEezsy8qwMrDyTz3rID7D6ZzZLYJJbuTaJnU3+e6BFKR2936PyA9sVg909waBkk7tAeoZ20XiS7Vbs/kVsgeIZo1x0JIUQNkpFnxu9kDAaDHbN/K0x+jR0dkhCimlLP3HrkYr1Mrq6uxYkQQEhICMnJyQCkpKQUT9DwwAMPFK9jtVrx8vIqsZ+zyUh5duzYQWhoaHEidD6bzcZbb73Fjz/+yMmTJykqKqKoqAg3twt/X3vkkUe45ZZb+PfffxkwYADDhg2je/fuF9zmSpNkSFwRfZoH0qd5ICv2J/PpysNsjstg9cFUVh9MpVO4D4/3qEcvf0/oOk67cevOH+DIcjixRXs06A5t7wC7DXISwdkLPOuBqx/IjdaEEDXAP/uTGaxoQ+RMbW9xcDRC1F4uRj17pw502LErokmTJiiKwt69exk2bFip5fv378fHxwd/f/8L7sdoLHkZgaIoxYmU/cwN7r/44gu6dOlSYr3zrzu6WNJy/lC+87333nt88MEHfPjhh7Ru3Ro3NzfGjx+P2Wy+4HaDBw/m+PHjLFy4kL///pt+/frx6KOPMm3atAtudyVJMiSuqL4tAunbIpCtcenM+Ocwqw6lsCUug9FxGUTV9+TR7sEMCvFE6fEktB4BO+dp9yWKX689/JtDs0EQ3lObitvgpF1X5B4MBpOjX54QQpRr3e4DvHN2iFzLYQ6NRYjaTFGUCg1VcyQ/Pz/69+/PzJkzeeqpp0okG0lJScydO5fRo0cX9wyZTCZstspdBxUUFET9+vU5evQod91112XF26ZNG06cOMHBgwfL7B1as2YNN910E3fffTegJWKHDh0qMQywvNcQEBDAmDFjGDNmDD179uSZZ55xaDIks8mJq6JjuC+z7+3Moid6MjgqGL1OYc/JbB6Zf5D+8zL5OSkQe2BL6DkRbpyhJT+KXptoYf10mD8aNn4Kyfsg47jWe5RyAAqzHf3ShBCilAKzDbejizEodgr8WoEMkROizpsxYwZFRUUMHDiQ1atXk5CQwJIlS+jfvz/169fn9ddfL143PDyc1atXc+rUKVJTUyt8jMmTJ/Pmm28WT2Cwe/duZs2axfvvv1+pWHv37k2vXr245ZZbiImJ4dixYyxevJglS5YAWk9XTEwM69evZ9++fTz00EMkJSWV2Ed4eDibNm0iLi6O1NRU7HY7L7/8Mn/88QeHDx8mNjaWBQsWlHsd1dUiyZC4qiJDPPn07mj+ebo3t0aH4mTQcTgll4l/HKbXt6l8G++POagN9H4Obp0NHcZovUCWfDiwEP56AhY+DQeXQEa8dvPWU9sh5zSc6R4WQghHW3s4lQHqBgCc28ksckIIaNq0KVu3bqVx48bcfvvtNG7cmAcffJC+ffuyYcMGfH19i9edOnUqx48fp0OHDgQFBVX4GPfffz9ffvkls2fPpnXr1vTu3ZvZs2cTERFR6Xh/+eUXOnXqxMiRI2nZsiXPPvtscU/PSy+9RIcOHRg4cCB9+vQhODi41PC/iRMnotfradmyJQEBAcTHx2MymXj++edp06YNvXr1Qq/X88MPP1Q6tqqkqGcHGlZj2dnZeHl5kZqaip+fn6PDqTYsFguLFi1iyJAhpcaQ1hSnswuYueII87edIN+svcECPJwY2zmEsa10uFgywGbRbtB6cAnEbwD1TJer0RUa9dGG0fk20qbjdg8CjxAs6Gt83VwJteGcuVKkbsqWlpaGv78/WVlZJWY0EhdumybPW8VL+29Cr6jwxHbtM6qOkPdS2aReyleZuiksLOTYsWNERETg7Ox8lSJ0DLvdTnZ2Np6enuh0da//4kJ/66psm6r3AEtR6wV5ujDlpigm9G/G/1YfZe6meFJyinhneRyfrNFzbfMAbo3ypWdET5R67aAgQ5to4eBSbWKFA4u0h3+zc9cWZZ8E45lZU+w2QBodIcTVY7XZ0R9cgF5RyfVthXsdSoSEEKKmkWRIVAteriaeHdSCx69twqz1ccxeF0dyThF/7U7ir90Q5OnEkJb+jGwZQLN2IdBqOCTt0pKi+A2QelB7bPkCGvWFxgMBH+3aInc/cPPXZqLTVWzWFyGEuFTbjmfQ17oO9OAiQ+SEEKJak2RIVCsuJgPj+jThoV6NWbH/ND9vO8nKg8mczi5i1saTzNoILYI9uKmVL7c2vRb/eh0gP/1Mb9GS4t4i44FFXOsUgs7cHRp0hYDmoDOAi48kRkKIK2rtzn2M1+0FQB8lN1oVQojqTJIhUS3pdQrXtQzmupbB5BdZ+WPnSX779xRbj6ezPymH/Uk5TFuh0CXchxGtwhnS9h6cW90MSbvh4BLU+A14FCVC7C/aw9kbwrpoiVFIWzA4n0uMXHxBL28FIcTlU1UV+96/0CsqWd5RePlW/qJlIYQQV498AxTVnquTgZGdGzKyc0NOZxXy09YE/th5isPJuaw/ms76o+m8aNLTr7kft7fsTfdrO2HLTWfH5tVEswfdya1QmAmHlmoPgzPUj4awrhDaCZw9JTESQlSJ/Uk5dC9cow2Ray83WhVCiOpOvvWJGiXIy5nH+zXl8X5N2ZeYxQ+bE1i0O4mU3CL+2p3MX7sh0MOJIS0C8DV2pU2PYThZ8+D0HojfCAkbIT8Njq/THooOgqKgQTet58gjGFy8tWF0zt5grN0z1Qghqtaa7fu478wQOVOb4Q6ORgghxMVIMiRqrMgQL6bc5MUrQ1ux5nAKP205wT/7k0nOKWL2ltOAgS8OZdKloRd9wnvRv1NPgrrmaRMtJGzUkqPM49pEDEm7YPP/tOlvw7pCWGfwiQAnd3D20hIjF29t+m4hhCiHec/v6BWVNK8o/HzCHR2OEEKIi5BkSNR4Op1C72aB9G4WSIHZxsLdifz+7wm2xKWSU2jl7wNp/H0AXgTC/Vy5Jrw114W3p2v7+3DOPq4lRgmbIHkvpB/VHju/B6MbBEZqPUdBrcCviZYQOXtpw+qcvWQSBiFEsRMZ+XTIWQl6cG4nQ+SEEKImkGRI1CouJj0jokO5qU0Qfy1YREDLLqw6lM6aw6kcTMohLi2fuLR85mwDJ4OO9qH16B1+F/173UljXRLKiS1aj1HSTrDkwcmt2gNA76TNShfUSkuQAluAW8C5niMnT6iDN0UTQmjWbo/lVt0+ANwkGRJCiBpBkiFRa+l10LWRHz2bBwOQnmdmxf5kVhxIZsORNNLyzGyMy2JjXBZvAwEeTnRvOIh+zYbS65oivHMOwulY7Xqj07FQlH1uSB2Aogf/pueSo6Ao8Kx3JjnyBJO79BwJUYfk79SGyCV7RhHo09DR4QghaqC4uDgiIiLYvn07bdq0qZJ9KorCb7/9xrBhw6pkf7WNJEOizvB1M3FLdCi3RIeiqioHT+fy997TrDyYzI6ETFJyzPyxJ50/9mjrh/k0ICqoOdH1biO6rUpL9QhOqXvOJEd7tIkYUvZrjz2/AAr4RkBgK60Hya+xNrTO2Vu79sjJQxt6J71HQtQ6GXlmWmb8AzowtZFeISFEaWPGjOGbb74pfu7r60unTp145513qizxqQ5WrlxJ3759ycjIwNvb29HhXJQkQ6JOUhSF5sEeNA/24NFrm1BgtrHpaBp/70tm7eEU4tLyScgoJCGjkMX7tW0MOm+aBPSnbdANREfr6OxyigZ5O9El79WSo+xT56452v+XtpHeBN4NtYkZfBtpCVJQS3AL+k+C5AqK4rjKEEJctnU7YhmiaEPkvDuOcHA0QojqatCgQcyaNQuApKQkXnzxRW644Qbi4+MdHFndJf+iFgLtWqM+LQJ57eYoVj7Tl+0v9Wf2mE482rcxPZr44e1ixGpX2X86nx93ZfBsTBp9/nQiak03RmQ8ymsNvmJFj7lkdHkOtfkNEBCp3c/IZoa0Q9r9jTZ9Cosmwqzr4duh8PN9sPj/YN10OLwcUg9DbjKY80FVHV0lQohKyNn+CzpFJdE9CrwbODocIUQ15eTkRHBwMMHBwbRr147nnnuOhIQEUlJSylzfZrNx//33ExERgYuLC82bN2f69Oml1vv6669p1aoVTk5OhISE8Nhjj5Ubw9SpUwkKCmLHjh3lrvPnn3/SsWNHnJ2d8ff3Z/jwc7cKmDNnDh07dsTDw4Pg4GDuvPNOkpOTAW2YX9++fQHw8fFBURTGjBkDwM8//0zr1q1xcXHBz8+P6667jry8vItV2RUnPUNClMHHzUSfFoH0aREIaHeVP5FRwI6ETLYez2BnQib7E7PJN9vYeiKPrSfy+BIFaIufW0ea+zvRrImOaNckWtoPU7/oCM7ZRyH9GBSka71I2afg+NpzB3X21obZ+USAVxj4hoN/c/CsDyY3MLmCwUWG2QlRzRSYbTRO+RsUUKJudnQ4QtQ9qgqWfMcc+zJGd+Tm5jJ37lyaNGmCn59fmevY7XZCQ0P56aef8Pf3Z/369Tz44IOEhIRw2223AfDpp58yYcIE3nrrLQYPHkxWVhbr1q0rtS9VVRk/fjy///47a9eupWnTpmUec+HChQwfPpxJkybx3XffYTabWbhwYfFys9nMq6++SvPmzUlOTuapp55izJgxLFq0iLCwMH755RduueUWDhw4gKenJy4uLiQmJjJy5Ejeeecdbr75ZnJyclizZg1qNfjnryRDQlSAoiiE+boS5uvK0Lb1ALDa7BxKzmVHfCbb4rUE6UhKLml5FtbnWVh/HGbjCrQB2uDjYqCRr4n2wTl0NB2jue0owUVHcc4+hpJ9Egoz4dR27fFfRhfwqKdNzuBZH3zCwbcxBDTTyowu2oexTNYghEPs2HeQIWjjaYO63ubgaISogyz58EY9xxz7hVPaPywraMGCBbi7uwOQl5dHSEgICxYsQFfOPzqNRiOTJ08uXh4REcH69ev56aefipOh1157jaeffponn3yyeLtOnTqV2I/VamX06NFs3bqVdevWERoaWm6Mr7/+OnfccQdTpkwpLmvbtm3x7/fee2/x740aNeKjjz6ic+fO5Obm4u7ujq+vLwCBgYHF1wwdOXIEq9XK8OHDadhQm2CmdevWF66sq0SSISEukUGvIzLEk8gQT0Z20YbF5JutxJ7K5mBSDgdO53DodC5HU3M5nV1ERoGVbSetbDup50uaAE0AcDXpiPRWucb9JO0Nx2hkPYav5RSu+afQ5aWApQDSj2iP8zl5nEuUvBtqPUu+jbRhOp71tUTJ4Ax6easLcaVk7vwTnaKS4NaKMBkiJ4S4gL59+/Lpp58CkJ6ezsyZMxk8eDCbN28uThLO99lnn/H1119z/PhxCgoKMJvNtGvXDoDk5GROnTpFv379Lnjcp556CicnJzZu3Ii/v/8F192xYwcPPPBAucu3b9/O5MmT2bFjB+np6djtdgDi4+Np2bJlmdu0bduWfv360bp1awYOHMiAAQMYMWIEPj4+F4zlapBvSEJUIVeTgU7hvnQK9y1Rnldk5UhKLoeTczl4OocDSTkcScnjZEYB+WY725JhW3I9oB7Qo3g7PycrHdzSaeN0mmaGJBqopwi0nMSzKBFjYRoU5UDRAUg9UDoYnQFc/cHNH9yDwCMYnXsIQVk2OOoK/o21cqMMvRPicoQl/wNOYGsxzNGhCFE3GV21HhpHHbsS3NzcaNKkSfHz6OhovLy8+OKLL3jttddKrf/bb7/x9NNP895779GtWzc8PDx499132bRpEwAuLi4VOm7//v2ZN28eS5cu5a677rrguhfaZ15eHgMGDGDAgAHMmTOHgIAA4uPjGThwIGazudzt9Ho9MTExrF+/nmXLlvHxxx8zadIkNm3aRERERIVew5UiyZAQV4Gbk4E2od60CfUuUW622olLy+Nwci6HTudw8ExP0qnMQrIKLKQVGYgpCiSGQKBkd7ILhTQzJNPBNYVIUxKNlUTq2U7hbU3G2ZyBYrdCbpL2OK3NF64HugIc/UDbicEJXAPAPQA8QrReJo9AcA/WHh5BWg+Ti4/0LglRjrbqIQBCe9zh4EiEqKMUpVJD1aoTRVHQ6XQUFBSUuXzDhg10796dcePGFZcdOXJupIiHhwfh4eEsX768eOKCstx4440MHTqUO++8E71ezx13lP951aZNG5YvX87YsWNLLdu/fz+pqam89dZbhIWFAbB169YS65hMJkCb/OH819qjRw969OjByy+/TMOGDfntt9+YMGFCubFcDfLtRggHMhl0NAvyoFmQB7QOKbEsr8jKycwCTmYUcCIjnxOZBSSkF3AiPZ+TWQWk5cJOawN2ZpcelmPAShAZhChphOrTaWzKIFyfTn1dKt6WFIKUdNxs2WAtguwT2oPtpfZTTG8CF28tKXL10x5uAdoU4e6B4BGsPTzrg6sv6I1VW1FCVGM6RSXOJYpwXxkiJ4S4sKKiIpKSkgDIyMhgxowZ5ObmMnTo0DLXb9SoET/++CNLly4lIiKC7777ji1btpToTZk8eTIPP/wwgYGBDB48mJycHNatW8fjjz9eYl8333wz3333HaNGjcJgMDBiRNm3AXjllVfo168fjRs35o477sBqtbJ48WKeffZZGjRogMlk4uOPP+bhhx9mz549vPrqqyW2b9iwIYqisGDBAoYMGYKLiwuxsbEsX76cAQMGEBgYyKZNm0hJSSEyMvJyqrNKSDIkRDXl5mQ4lyiVodBi41RmwX8SJi1pSswqJCWniNRcZ04WBrDVClhLb++EmRAljRAlnXr/3969x0ZRt3sA/87O3qEXoUBbtX1LwnmL4AWo5nAtRqkHCN4VUS4B5YQASulRKWCCYKDITYLlYg2RGIIS4g0SUBolrQgGLFQ5oBKPDeXFktK+2t12u7uzO7/zx+y2XTqLLZTOlv1+ksnM/HaWPn3szuOzs/sb1CNNqsed8r+RJjcgRXKhLxqQLBpgE15tivDGWm3pCItT+z6TLRGwJ2mLIxmw36Y1S44+2tqZAvTqGxq7DbD25j2XqEfy/sejRodARD3Al19+ibQ07c3PhIQEZGdnY9++fRg/frzu8bNnz8Yvv/yCqVOnQpIkTJs2DfPnz8ehQ4dajpk1axa8Xi/eeecdvPrqq0hJSYna6Dz99NNQVRUzZsyAyWSKmDI7bPz48di3bx/eeustrF27FomJiRg3bhwAoF+/fti1axeWLVuGLVu2YPjw4diwYQMefbT1HHj77bdj5cqVKCwsxOzZszFz5kwsWbIE5eXl2Lx5M1wuFzIzM7Fx40ZMnDjxelPZZSRxHXPabdu2DevXr0dNTQ2GDBmCzZs3Y+zYsVGPLysrQ0FBAc6ePYv09HS8/vrrmDdvXod/nsvlQlJSEurq6qJOPRiPFEXBwYMHMWnSJFgsfCe+LeZG4wsEUd/oxxW3D1fcPlxu8ODYqTO4Le0fqPcoqHV5caVRe8yrqLr/hgNe9JVc6IcG9JVcSJEa0Bfaul9ou79JeywZ7hsLWDJpjZS1V2jprS223lpjZUvQFntSZKPVcgPb0BTkFoe2bbZ2+Efzb0ZffX09UlJS0NDQgMTERKPDiSnh2tRQmADL//wvHCm8MgTwtRQN8xJdZ3Lj9XpRVVWFrKws2O32borQGKqqwuVyITExMepsc7eya/237sra1OkrQ3v37kV+fj62bduG0aNH47333sPEiRNx7tw5ZGS0LwRVVVWYNGkS5s6di927d+O7777D/Pnz0a9fPzz11FM3FDwRXZvNLCM92YH0ZO3LkIqiIKH2J0yaNLhdwWnyBXDF7cNfzQr+9PjR4FHwl8ePv5oV/NVm+2ePP7SvwNWsoO27KTKCSIAHyVIjktCEZKkJSWhEktSE5PBaakISmpAkNbaOoQk2SQGECvgbtaULCEmGMNshzI5Qg+SAZO0FKbTWGiet8TLJdvyz5l8wHf9Na8TMNm0mPotdW4f3zVftWxzatmzTpjfnla2483/Wf2IYGyEioh6p083Qpk2b8OKLL+Kll14CAGzevBlfffUVtm/fjqKionbH79ixAxkZGdi8eTMAYPDgwfjhhx+wYcMGNkNEMaSXzYxets6dElRVwO0N4K/mUIPUrKDJF0CjNwB3aN3oU9DoC6DKG0CjL4AmXwDu0Lbbq+0HVAE7fEiAB70lL3qjGb2lZm3dsu1Fb8kT2ve2PNZLakZC6BgHfHDCB7OkXeWSRBCS0gQoTYD+d1NbyACyAeDyZ9eVPwAQkBA0WaBKFqgmK4TJAtVkgZAtECYrhKwtkC0Qsk37LpbZBkm2QpItgGyBJJtD69bFJFshmS0wyRZIZmtobYEkW7VZA2WLtpZkbW0Kr8OL6ar90DFtj5dk7ThJbrMvaw0qXZP7H/9ldAhERHSdOvV/Pn6/HxUVFSgsLIwYz8vLw7Fjx3Sfc/z4ceTl5UWMPfLII9i5cycUReGlYqIezGSSkOS0IMlpQeZ1foJVCAFfQIXbG4DHH0CzEoTHH4TXr62blSCa/cHQYyrq/QH8S4l8LPwcX0CFTwlCVXyQAh5ISjPkYDPkQDMsqhdOyQcHQovkh7NlW2uirFBggwK75IcttG2TFNih7Vuv2rfBD6vUOluOBAGz6gfgB4JNXZNkg1l8xt8dPNbd8Z/tP3NPREQ9Q6eaobq6OgSDQQwYMCBifMCAAS0zY1zt8uXLuscHAgHU1dW1fImsLZ/PB5/P17LvcrkAaB/xURSlMyHf0sK5YE7aY270xWpeZADJdhOS7R3/jk9nqaqAP6hqDVNAhVfRmid/aL/J68eJHypw9z33QZVMcAdU/DuoQgkKKEE1tGj/htJmPBBQIBQfEPBBBBVIQT8Q9EEK+iGpSsvapPohBRWYQtsmVYEstG1ZKJBFACYRgCyCbbYDMCMIM4KwSEGYoe1bwmOhfbMUhAwVMoIwh9bavgozgpCl0LrN463rIGSJDU9HRKtNztvSYu41ZaRYPc8YjXmJrjO5URQFQgioqtpys89bVfhr/eHfN96oqgohBBRFgSzLEY915evoumaTk676TLwQot3Y3x2vNx5WVFSElStXths/cuQInM7O3dwqHpSWlhodQsxibvQxL/oGJwOBam2KcXNo+dvb2YUPjCCHntmxm+FdixCACkAVgF8AXqFth8fCi2i7H3peeKztvyEACCFFPEeI0EFQAVUFoEKCCklV0expAvDfN/x73ApYmzqH5xl9zEt0HcmN2WxGamoqGhsbr3mTz1uJ232DkxP1UH6/H83NzSgvL0cgEDktrsfj6bKf06lmKCUlBbIst7sKVFtb2+7qT1hqaqru8WazOerMcEuXLo24AZPL5cKdd96JBx98kLPJtaEoCkpLSzFhwgR+3PAqzI0+5iU65kZffX09Vi02OorYwNrUMXwt6WNeoutMbnw+H6qrq+F0Om/5NyGEEHC73UhISLjmRYdblcfjgcPhQG5uLmw2W8Rj9fX1XfZzOtUMWa1WjBgxAqWlpXjiiSdaxktLS/HYY4/pPmfkyJE4cOBAxNjhw4eRk5MT9Q/eZrO1+6UBwGKx8ASig3mJjrnRx7xEx9xEYi5asTZ1DvOij3mJriO5kWW55Y35fv36wWq13rKNgqqq8Pv98Pl8cTW1thACfr8fV65cgSzLcDqd7X7/rnwNdfpjcgUFBZgxYwZycnIwcuRIlJSUoLq6uuW+QUuXLsWlS5fw4YcfAgDmzZuH4uJiFBQUYO7cuTh+/Dh27tyJjz76qMt+CSIiIiK69ZlMJmRlZaGmpgZ//PGH0eHcVEIINDc3w+Fw3LIN37U4nU5kZGTc9Eaw083Q1KlTtY9NrFqFmpoaDB06FAcPHkRmZiYAoKamBtXV1S3HZ2Vl4eDBg1i8eDG2bt2K9PR0bNmyhdNqExEREVGnWa1WZGRkIBAIIBgM/v0TeihFUVBeXo5x48bF3dVEWZZhNpu7pQm8rgkU5s+fj/nz5+s+tmvXrnZjubm5OHXq1PX8KCIiIiKiCJIk3fIfOZRlGYFAAHa7/Zb+PY0WPx9AJCIiIiIiaoPNEBERERERxSU2Q0REREREFJeu6ztD3S18k1a3283PTLahKAo8Hg9cLhfzchXmRh/zEh1zoy98s7/weZhasTbp42tJH/MSHXOjj3mJritrU49ohsI3VsrKyjI4EiKi+FRfX4+kpCSjw4gprE1ERMbqitrUI5qhPn36AACqq6tZjNsI3/384sWLSExMNDqcmMLc6GNeomNu9DU0NCAjI6PlPEytWJv08bWkj3mJjrnRx7xE15W1qUc0Q+GbLSUlJfGPQUdiYiLzEgVzo495iY650RdPdz/vKNama+NrSR/zEh1zo495ia4rahOrGxERERERxSU2Q0REREREFJd6RDNks9mwYsUK2Gw2o0OJKcxLdMyNPuYlOuZGH/MSHXOjj3nRx7xEx9zoY16i68rcSILzpRIRERERURzqEVeGiIiIiIiIuhqbISIiIiIiiktshoiIiIiIKC6xGSIiIiIiorgU883Qtm3bkJWVBbvdjhEjRuDbb781OiTDFRUV4f7770dCQgL69++Pxx9/HL/++qvRYcWcoqIiSJKE/Px8o0OJCZcuXcL06dPRt29fOJ1O3HfffaioqDA6LEMFAgG88cYbyMrKgsPhwMCBA7Fq1Sqoqmp0aN2uvLwcU6ZMQXp6OiRJwueffx7xuBACb775JtLT0+FwODB+/HicPXvWmGBjAGtTe6xNHcPaFIm1qT3WplbdUZtiuhnau3cv8vPzsXz5cpw+fRpjx47FxIkTUV1dbXRohiorK8OCBQvw/fffo7S0FIFAAHl5eWhqajI6tJhx8uRJlJSU4J577jE6lJjw559/YvTo0bBYLDh06BDOnTuHjRs3Ijk52ejQDPX2229jx44dKC4uxs8//4x169Zh/fr1ePfdd40Ords1NTXh3nvvRXFxse7j69atw6ZNm1BcXIyTJ08iNTUVEyZMgNvt7uZIjcfapI+16e+xNkVibdLH2tSqW2qTiGEPPPCAmDdvXsRYdna2KCwsNCii2FRbWysAiLKyMqNDiQlut1sMGjRIlJaWitzcXLFo0SKjQzLckiVLxJgxY4wOI+ZMnjxZzJkzJ2LsySefFNOnTzcootgAQHz22Wct+6qqitTUVLF27dqWMa/XK5KSksSOHTsMiNBYrE0dw9oUibWpPdYmfaxN+m5WbYrZK0N+vx8VFRXIy8uLGM/Ly8OxY8cMiio2NTQ0AAD69OljcCSxYcGCBZg8eTIefvhho0OJGfv370dOTg6eeeYZ9O/fH8OGDcP7779vdFiGGzNmDL7++mucP38eAPDjjz/i6NGjmDRpksGRxZaqqipcvnw54nxss9mQm5sbd+dj1qaOY22KxNrUHmuTPtamjumq2mS+GcF1hbq6OgSDQQwYMCBifMCAAbh8+bJBUcUeIQQKCgowZswYDB061OhwDPfxxx/j1KlTOHnypNGhxJTff/8d27dvR0FBAZYtW4YTJ07glVdegc1mw8yZM40OzzBLlixBQ0MDsrOzIcsygsEgVq9ejWnTphkdWkwJn3P1zscXLlwwIiTDsDZ1DGtTJNYmfaxN+libOqaralPMNkNhkiRF7Ash2o3Fs4ULF+Knn37C0aNHjQ7FcBcvXsSiRYtw+PBh2O12o8OJKaqqIicnB2vWrAEADBs2DGfPnsX27dvjuuDs3bsXu3fvxp49ezBkyBBUVlYiPz8f6enpmDVrltHhxRyej1sxF9fG2tSKtSk61iZ9rE2dc6Pn45hthlJSUiDLcrt32mpra9t1gPHq5Zdfxv79+1FeXo477rjD6HAMV1FRgdraWowYMaJlLBgMory8HMXFxfD5fJBl2cAIjZOWloa77rorYmzw4MH45JNPDIooNrz22msoLCzEc889BwC4++67ceHCBRQVFbHgtJGamgpAexcuLS2tZTwez8esTX+PtSkSa1N0rE36WJs6pqtqU8x+Z8hqtWLEiBEoLS2NGC8tLcWoUaMMiio2CCGwcOFCfPrpp/jmm2+QlZVldEgx4aGHHsKZM2dQWVnZsuTk5OCFF15AZWVl3BYbABg9enS7KW7Pnz+PzMxMgyKKDR6PByZT5GlQluW4nL70WrKyspCamhpxPvb7/SgrK4u78zFrU3SsTfpYm6JjbdLH2tQxXVWbYvbKEAAUFBRgxowZyMnJwciRI1FSUoLq6mrMmzfP6NAMtWDBAuzZswdffPEFEhISWt6hTEpKgsPhMDg64yQkJLT7bHqvXr3Qt2/fuP/M+uLFizFq1CisWbMGzz77LE6cOIGSkhKUlJQYHZqhpkyZgtWrVyMjIwNDhgzB6dOnsWnTJsyZM8fo0LpdY2Mjfvvtt5b9qqoqVFZWok+fPsjIyEB+fj7WrFmDQYMGYdCgQVizZg2cTieef/55A6M2BmuTPtYmfaxN0bE26WNtatUttakrprq7mbZu3SoyMzOF1WoVw4cP5xSdQptaUG/54IMPjA4t5nD60lYHDhwQQ4cOFTabTWRnZ4uSkhKjQzKcy+USixYtEhkZGcJut4uBAweK5cuXC5/PZ3Ro3e7IkSO655VZs2YJIbQpTFesWCFSU1OFzWYT48aNE2fOnDE2aAOxNrXH2tRxrE2tWJvaY21q1R21SRJCiBtq2YiIiIiIiHqgmP3OEBERERER0c3EZoiIiIiIiOISmyEiIiIiIopLbIaIiIiIiCgusRkiIiIiIqK4xGaIiIiIiIjiEpshIiIiIiKKS2yGiIiIiIgoLrEZIiIiIiKiuMRmiIiIiIiI4hKbISIiIiIiiktshoiIiIiIKC79P8L7MB7WjRrQAAAAAElFTkSuQmCC", "text/plain": [ "
" ] @@ -1061,6 +1003,7 @@ "for key, value in lambdas.items():\n", " lower, median, upper = value\n", " ax[0].plot(t, S0(median * t), label=f\"{key}\")\n", + " ax[0].fill_between(t, S0(lower * t), S0(upper * t), alpha=0.25)\n", "\n", "ax[0].grid(True)\n", "ax[0].set_xlim(0, 10)\n", @@ -1070,6 +1013,7 @@ "for key, value in lambdas.items():\n", " lower, median, upper = value\n", " ax[1].plot(t, cdf(median * t), label=f\"{key}\")\n", + " ax[1].fill_between(t, cdf(lower * t), cdf(upper * t), alpha=0.25)\n", "\n", "ax[1].grid(True)\n", "ax[1].set_xlim(0, 10)\n", @@ -1086,7 +1030,7 @@ }, { "cell_type": "code", - "execution_count": 233, + "execution_count": null, "metadata": {}, "outputs": [ { @@ -1124,7 +1068,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Scaling adoption times back to days, we can see that black cats have longer and a wider range of time until adoptions (about 55 days) than cats that are not black (about 51 days)." + "Scaling adoption times back to days (multiplying by 31), we can see that black cats have longer and a wider range of time until adoptions (about 55 days) than cats that are not black (about 51 days)." ] }, { @@ -1138,7 +1082,7 @@ }, { "cell_type": "code", - "execution_count": 36, + "execution_count": null, "metadata": {}, "outputs": [ {