Update file calls in calibration modules

.gitignore

@@ -24,7 +24,6 @@ htmlcov/*
 *.asv
 *.nav
 *.snm
-*.gz
 *.bib.bak
 *.fls
 *.m~
@@ -47,6 +46,7 @@ examples/OG-USA-Example/*
 cs-config/cs_config/OUTPUT_BASELINE/*
 data/csv_output_files/*
 data/images/*
+data/PSID/psid_lifetime_income.csv
 ogusa/csv_output_files/*
 ogusa/images/*
 .vscode/

CHANGELOG.md

@@ -5,6 +5,14 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+
+## [0.1.8] - 2024-05-20 12:00:00
+
+### Added
+
+- Updates the file calls in some of the calibration modules (`bequest_transmission.py`,  `deterministic_profiles.py`, `psid_data_setup.py`, and `transfer_distirbution.py`). These files called `.RData` files that were not included in the `ogusa` package, so users who pip installed ogusa could not instantiate the `Calibrate` class object. These changes allow for Jupyter notebook users to automatically download the data files from the master branch of the GitHub repository if they are not working in their own branch of the repository.
+
+
 ## [0.1.7] - 2024-05-14 16:30:00
 
 ### Added
@@ -97,6 +105,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 
 
+[0.1.8]: https://github.com/PSLmodels/OG-USA/compare/v0.1.7...v0.1.8
 [0.1.7]: https://github.com/PSLmodels/OG-USA/compare/v0.1.6...v0.1.7
 [0.1.6]: https://github.com/PSLmodels/OG-USA/compare/v0.1.5...v0.1.6
 [0.1.5]: https://github.com/PSLmodels/OG-USA/compare/v0.1.4...v0.1.5

ogusa/bequest_transmission.py

@@ -3,42 +3,40 @@
 import matplotlib.pyplot as plt
 import os
 from ogusa.utils import MVKDE
-
-CURDIR = os.path.split(os.path.abspath(__file__))[0]
+from ogusa.constants import CODE_PATH
 
 
 def get_bequest_matrix(
     J=7,
     lambdas=np.array([0.25, 0.25, 0.2, 0.1, 0.1, 0.09, 0.01]),
-    graphs=False,
+    data_path=None,
+    output_path=None,
 ):
     """
     Returns S x J matrix representing the fraction of aggregate
     bequests that go to each household by age and lifetime income group.
 
+    Args:
+        J (int): number of lifetime income groups
+        lambdas (Numpy array): length J array of lifetime income group
+            proportions
+        data_path (str): path to PSID data
+        output_path (str): path to save output plots and data
+
+    Returns:
+        kde_matrix (Numpy array): SxJ shaped array that represents the
+            smoothed distribution of proportions going to each (s,j)
+
     """
-    # Create directory if output directory does not already exist
-    CURDIR = os.path.split(os.path.abspath(__file__))[0]
-    output_fldr = "csv_output_files"
-    output_dir = os.path.join(CURDIR, "..", "data", output_fldr)
-    if not os.access(output_dir, os.F_OK):
-        os.makedirs(output_dir)
-    image_fldr = "images"
-    image_dir = os.path.join(CURDIR, "..", "data", image_fldr)
-    if not os.access(image_dir, os.F_OK):
-        os.makedirs(image_dir)
-
-    # Define a lambda function to compute the weighted mean:
-    # wm = lambda x: np.average(
-    #     x, weights=df.loc[x.index, "fam_smpl_wgt_core"])
-
-    # Read in dataframe of PSID data
-    # df = ogcore.utils.safe_read_pickle(
-    #     os.path.join(CURDIR, "data", "PSID", "psid_lifetime_income.pkl")
-    # )
-    df = pd.read_csv(
-        os.path.join(CURDIR, "..", "data", "PSID", "psid_lifetime_income.csv")
-    )
+    # Read in PSID data
+    if data_path is None:
+        # Read data file shipped with OG-USA package
+        df = pd.read_csv(
+            os.path.join(CODE_PATH, "psid_lifetime_income.csv.gz")
+        )
+    else:
+        # This is the case when running this from a branch of the OG-USA repo
+        df = pd.read_csv(data_path)
 
     # Do some tabs with data file...
     # 'net_wealth', 'inheritance', 'value_inheritance_1st',
@@ -54,12 +52,15 @@ def get_bequest_matrix(
     )
     # print(df[['sum_inherit', 'inheritance']].describe())
 
-    if graphs:
+    if output_path is not None:
+        # Create plot path directory if it doesn't already exist
+        if not os.path.exists(output_path):
+            os.makedirs(output_path)
         # Total inheritances by year
         df.groupby("year_data").mean(numeric_only=True).plot(y="inheritance")
-        plt.savefig(os.path.join(image_dir, "inheritance_year.png"))
+        plt.savefig(os.path.join(output_path, "inheritance_year.png"))
         df.groupby("year_data").mean(numeric_only=True).plot(y="sum_inherit")
-        plt.savefig(os.path.join(image_dir, "sum_inherit_year.png"))
+        plt.savefig(os.path.join(output_path, "sum_inherit_year.png"))
         # not that summing up inheritances gives a much larger value than
         # taking the inheritance variable
 
@@ -68,22 +69,22 @@ def get_bequest_matrix(
         df[df["year_data"] >= 1988].groupby("age").mean(
             numeric_only=True
         ).plot(y="net_wealth")
-        plt.savefig(os.path.join(image_dir, "net_wealth_age.png"))
+        plt.savefig(os.path.join(output_path, "net_wealth_age.png"))
         df[df["year_data"] >= 1988].groupby("age").mean(
             numeric_only=True
         ).plot(y="inheritance")
-        plt.savefig(os.path.join(image_dir, "inheritance_age.png"))
+        plt.savefig(os.path.join(output_path, "inheritance_age.png"))
 
         # Inheritances by lifetime income group
         # bar plot
         df[df["year_data"] >= 1988].groupby("li_group").mean(
             numeric_only=True
         ).plot.bar(y="net_wealth")
-        plt.savefig(os.path.join(image_dir, "net_wealth_li.png"))
+        plt.savefig(os.path.join(output_path, "net_wealth_li.png"))
         df[df["year_data"] >= 1988].groupby("li_group").mean(
             numeric_only=True
         ).plot.bar(y="inheritance")
-        plt.savefig(os.path.join(image_dir, "inheritance_li.png"))
+        plt.savefig(os.path.join(output_path, "inheritance_li.png"))
 
         # lifecycle plots with line for each ability type
         pd.pivot_table(
@@ -93,15 +94,15 @@ def get_bequest_matrix(
             columns="li_group",
             aggfunc="mean",
         ).plot(legend=True)
-        plt.savefig(os.path.join(image_dir, "net_wealth_age_li.png"))
+        plt.savefig(os.path.join(output_path, "net_wealth_age_li.png"))
         pd.pivot_table(
             df[df["year_data"] >= 1988],
             values="inheritance",
             index="age",
             columns="li_group",
             aggfunc="mean",
         ).plot(legend=True)
-        plt.savefig(os.path.join(image_dir, "inheritance_age_li.png"))
+        plt.savefig(os.path.join(output_path, "inheritance_age_li.png"))
 
     # Matrix Fraction of inheritances in a year by age and lifetime_inc
     inheritance_matrix = pd.pivot_table(
@@ -114,16 +115,18 @@ def get_bequest_matrix(
     # replace NaN with zero
     inheritance_matrix.fillna(value=0, inplace=True)
     inheritance_matrix = inheritance_matrix / inheritance_matrix.sum().sum()
-    # inheritance_matrix.to_csv(os.path.join(
-    #     output_dir, 'bequest_matrix.csv'))
 
     # estimate kernel density of bequests
+    if output_path is not None:
+        filename = os.path.join(output_path, "inheritance_kde.png")
+    else:
+        filename = None
     kde_matrix = MVKDE(
         80,
         7,
         inheritance_matrix.to_numpy(),
-        filename=os.path.join(image_dir, "inheritance_kde.png"),
-        plot=graphs,
+        filename=filename,
+        plot=(output_path is not None),
         bandwidth=0.5,
     )
 
@@ -139,10 +142,11 @@ def get_bequest_matrix(
         )
         kde_matrix = kde_matrix_new
 
-    np.savetxt(
-        os.path.join(output_dir, "bequest_matrix_kde.csv"),
-        kde_matrix,
-        delimiter=",",
-    )
+    if output_path is not None:
+        np.savetxt(
+            os.path.join(output_path, "bequest_matrix_kde.csv"),
+            kde_matrix,
+            delimiter=",",
+        )
 
     return kde_matrix

ogusa/calibrate.py

@@ -1,6 +1,6 @@
 from ogusa import estimate_beta_j, bequest_transmission
 from ogusa import macro_params, transfer_distribution, income
-from ogusa import get_micro_data, psid_data_setup
+from ogusa import get_micro_data
 import os
 import numpy as np
 from ogcore import txfunc, demographics
@@ -25,6 +25,8 @@ def __init__(
         data="cps",
         client=None,
         num_workers=1,
+        demographic_data_path=None,
+        output_path=None,
     ):
         """
         Constructor for the Calibration class.  This class is used to find
@@ -43,10 +45,15 @@ def __init__(
             data (str): data source for microsimulation model
             client (Dask client object): client
             num_workers (int): number of workers for Dask client
+            output_path (str): path to save output to
 
         Returns:
             Calibration class object instance
         """
+        # Create output_path if it doesn't exist
+        if output_path is not None:
+            if not os.path.exists(output_path):
+                os.makedirs(output_path)
         self.estimate_tax_functions = estimate_tax_functions
         self.estimate_beta = estimate_beta
         self.estimate_chi_n = estimate_chi_n
@@ -76,10 +83,14 @@ def __init__(
         self.macro_params = macro_params.get_macro_params()
 
         # eta estimation
-        self.eta = transfer_distribution.get_transfer_matrix(p.J, p.lambdas)
+        self.eta = transfer_distribution.get_transfer_matrix(
+            p.J, p.lambdas, output_path=output_path
+        )
 
         # zeta estimation
-        self.zeta = bequest_transmission.get_bequest_matrix(p.J, p.lambdas)
+        self.zeta = bequest_transmission.get_bequest_matrix(
+            p.J, p.lambdas, output_path=output_path
+        )
 
         # demographics
         if estimate_pop:
@@ -92,6 +103,7 @@ def __init__(
                 initial_data_year=p.start_year - 1,
                 final_data_year=p.start_year,
                 GraphDiag=False,
+                download_path=demographic_data_path,
             )
 
             # demographics for 80 period lives (needed for getting e below)
@@ -112,15 +124,15 @@ def __init__(
                 self.demographic_params["omega_SS"],
                 demog80["omega_SS"],
                 p.lambdas,
-                plot=False,
+                plot_path=output_path,
             )
         else:
             self.e = income.get_e_interp(
                 p.S,
                 p.omega_SS,
                 p.omega_SS,
                 p.lambdas,
-                plot=False,
+                plot_path=output_path,
             )
 
     # Tax Functions

ogusa/constants.py

@@ -1,10 +1,13 @@
 import taxcalc
+import os
 
 SHOW_RUNTIME = False  # Flag to display RuntimeWarnings when run model
 
 REFORM_DIR = "OUTPUT_REFORM"
 BASELINE_DIR = "OUTPUT_BASELINE"
 
+CODE_PATH = os.path.abspath(os.path.dirname(__file__))
+
 # Default year for model runs
 DEFAULT_START_YEAR = 2021
 # Tax-Calculator start year