added chart and optimization

pages/Micro_Finance_Analyzer.py

@@ -5,21 +5,46 @@
 """
 
 import logging
-import streamlit as st
-import pandas as pd
+from pathlib import Path
 
-# import sys
-# import os
+import numpy as np
+import pandas as pd
+import polars as pl
+import streamlit as st
+from scipy.stats import norm
 
-# # Add the src directory to the sys.path
-# sys.path.append(os.path.abspath(os.path.join(os.getcwd(), "src")))
-# import optimization as opt
+import src.charts as charts
 import src.optimization as opt
 
+st.set_page_config(
+    page_title="SEB Micro Finance Analyser",
+    layout="wide",
+    page_icon="💸",
+)
 
 logger: logging.Logger = logging.getLogger(__name__)
 
 
+def load_data() -> None:
+    if "corr" in st.session_state:
+        logger.debug("ignoring load_data because data already exist")
+        return
+
+    logger.info("Reading in data and calculating params")
+    rets = pd.read_csv("data/micfin_q_rets.csv", index_col=0, parse_dates=True)
+    st.session_state["orig_rets"] = rets
+
+    exp_rets = pd.read_csv("data/micfin_exp_rets.csv", index_col=0).squeeze()
+    st.session_state["orig_exp_rets"] = exp_rets
+
+    vols = rets.std() * 2  # * sqrt(4) for quarterly
+    vols.name = "Vol"
+    st.session_state["orig_vols"] = vols
+
+    corr = rets.corr()
+    st.session_state["orig_corr"] = corr.round(2)
+
+
 def get_user_input() -> None:
     with st.sidebar:
         st.title("Parameters")
@@ -37,6 +62,15 @@ def get_user_input() -> None:
         )
         st.session_state["r_min"] = r_min
 
+        all_tickers = st.session_state["orig_corr"].columns.tolist()
+        tickers = st.multiselect(
+            label="Tickers",
+            options=all_tickers,
+            default=all_tickers[:-1],
+            help="Select Tickers for the optimization",
+        )
+        st.session_state["tickers"] = tickers
+
         with st.popover("Exp Returns"):
             exp_rets = st.data_editor(
                 st.session_state["orig_exp_rets"] * 100,
@@ -75,33 +109,49 @@ def get_user_input() -> None:
             st.session_state["corr"] = corr.div(100)
 
 
-def load_data() -> None:
-    if "corr" in st.session_state:
-        logger.info("ignoring load_data because data already exist")
-        return
-
-    logger.info("Reading in data and calculating params")
-    rets = pd.read_csv("data/micfin_q_rets.csv", index_col=0, parse_dates=True)
-    st.session_state["orig_rets"] = rets
-
-    exp_rets = pd.read_csv("data/micfin_exp_rets.csv", index_col=0).squeeze()
-    st.session_state["orig_exp_rets"] = exp_rets
-
-    vols = rets.std() * 2  # * sqrt(4) for quarterly
-    vols.name = "Vol"
-    st.session_state["orig_vols"] = vols
-
-    corr = rets.corr()
-    st.session_state["orig_corr"] = corr.round(2)
-
-
 def main() -> None:
     st.title("Micro Finance Analyzer")
-    load_data()
+    load_data()  # Just once
+
     get_user_input()
-    st.dataframe(st.session_state["corr"])
-    st.dataframe(st.session_state["exp_rets"])
-    st.dataframe(st.session_state["vols"])
+
+    # get covar and calculate optimal portfolio
+    tickers = st.session_state["tickers"]
+    vols = st.session_state["vols"].loc[tickers]
+    corr = st.session_state["corr"].loc[tickers, tickers]
+    vol_d = np.diag(vols.values)
+    cov = vol_d @ corr.values @ vol_d
+    exp_rets = st.session_state["exp_rets"].loc[tickers]
+    r_min = st.session_state["r_min"]
+    w, r_opt, vol_opt = opt.find_min_var_portfolio(exp_rets, cov, r_min)
+
+    # Create Scatter chart and Portfolio Composition Chart
+    g_data = pl.DataFrame(
+        {
+            "name": tickers + ["OPTIMAL"],
+            "vols": vols.to_list() + [vol_opt],
+            "rets": exp_rets.to_list() + [r_opt],
+            "w_opt": w.tolist() + [1],
+        }
+    )
+    f_sc = charts.create_scatter_chart(g_data, None)
+    f_w = charts.create_portf_weights_chart(g_data)
+    title = f"Optimal Portfolio: r={r_opt:0.1%}, vol={vol_opt:0.1%}"
+    col1, col2 = st.columns([1.5, 1])
+    col1.altair_chart(f_sc, use_container_width=True)
+    col2.altair_chart(f_w.properties(title=title, height=350), use_container_width=True)
+    st.markdown(
+        f"""
+        ### Optimal Portfolio Statistics
+        * **Expected Return in 1y** = {r_opt:0.1%}
+        * **Expected volatility** = {vol_opt:0.1%}
+        * 95%-prob Lowest Expected Return in 1y = {(r_opt - norm.ppf(0.95)*vol_opt):0.1%}
+        * 99%-prob Lowest Expected Return in 1y = {(r_opt - norm.ppf(0.99)*vol_opt):0.1%}
+        """
+    )
+
+    st.divider()
+    st.caption(Path("data/disclaimer.txt").read_text())
 
 
 # Entry point for the script