refactor: logistic model

app.py

@@ -3,14 +3,18 @@ from typing import OrderedDict
 
 
 from src.features.build_features import initialise_data
+
 from src.models.xgboost_model import xgboost_class
+from src.models.logistic_model import logistic_class
+
+
 from src.models.util_strategy_table import strategy_table_view
 
 
 def main():
     currency_options = ["USD", "KES", "GBP"]
 
-    model_options = ["XGBoost"]
+    model_options = ["XGBoost", "Logistic"]
 
     currency = st.sidebar.selectbox(
         label="What currency will you be using?", options=currency_options
@@ -32,6 +36,10 @@ def main():
 
     model_classes = OrderedDict()
 
+    if "Logistic" in models_selected_set:
+        logistic_model_class = logistic_class(split_dataset, currency)
+        model_classes["Logistic"] = logistic_model_class
+
     if "XGBoost" in models_selected_set:
         xgboost_model_class = xgboost_class(split_dataset, currency)
         model_classes["XGBoost"] = xgboost_model_class

src/models/logistic_model.py

@@ -0,0 +1,33 @@
+from src.features.build_features import SplitDataset
+
+from src.models.logistic_train_model import logistic_train_model
+from src.models.logistic_predict_model import logistic_predict_model
+from src.models.logistic_test_model import logistic_test_model
+
+from src.models.util_model_class import ModelClass
+
+
+def logistic_class(split_dataset: SplitDataset, currency: str) -> ModelClass:
+
+    # Train Model
+    clf_logistic_model = logistic_train_model(split_dataset)
+
+    # Predict using Trained Model
+    clf_logistic_predictions = logistic_predict_model(
+        clf_logistic_model, split_dataset)
+
+    # Test and Evaluate Model
+    df_trueStatus_probabilityDefault_threshStatus_loanAmount_logistic = logistic_test_model(
+        clf_logistic_model,
+        split_dataset,
+        currency,
+        clf_logistic_predictions.probability_threshold_selected,
+        clf_logistic_predictions.predicted_default_status)
+
+    return ModelClass(
+        model=clf_logistic_model,
+        trueStatus_probabilityDefault_threshStatus_loanAmount_df=df_trueStatus_probabilityDefault_threshStatus_loanAmount_logistic,
+        probability_threshold_selected=clf_logistic_predictions.probability_threshold_selected,
+        predicted_default_status=clf_logistic_predictions.predicted_default_status,
+        prediction_probability_df=clf_logistic_predictions.prediction_probability_df,
+    )

src/models/logistic_predict_model.py

@@ -0,0 +1,4 @@
+from src.models.util_predict_model import make_prediction_view
+
+logistic_predict_model = make_prediction_view(
+    "Logistic", "Logisitic Model")

src/models/logistic_test_model.py

@@ -0,0 +1,4 @@
+from src.models.util_test import make_tests_view
+
+logistic_test_model = make_tests_view(
+    "Logistic", "Logistic Model")

src/models/logistic_train_model.py

@@ -0,0 +1,71 @@
+
+import numpy as np
+from sklearn.linear_model import LogisticRegression
+from src.features.build_features import SplitDataset
+import streamlit as st
+import pandas as pd
+
+from src.visualization.graphs_logistic import plot_logistic_coeff_barh
+
+
+@st.cache(suppress_st_warning=True)
+def create_clf_logistic_model(X_train, y_train):
+    # Create and fit the logistic regression model
+    return LogisticRegression(solver="lbfgs").fit(X_train, np.ravel(y_train))
+
+
+@st.cache(suppress_st_warning=True)
+def create_coeff_dict_logistic_model(
+    logistic_model, training_data
+):
+    return {
+        feat: coef
+        for coef, feat in zip(
+            logistic_model.coef_[0, :], training_data.columns
+        )
+    }
+
+
+def coeff_dict_to_sorted_df(coef_dict):
+    coef_dict_sorted = dict(
+        sorted(coef_dict.items(), key=lambda item: item[1], reverse=False)
+    )
+
+    data_items = coef_dict_sorted.items()
+    data_list = list(data_items)
+
+    return pd.DataFrame(data_list, columns=["Coefficient", "Value"])
+
+
+def interpret_clf_logistic_model(clf_logistic_model, split_dataset):
+    st.metric(
+        label="# of Coefficients in Logistic Regression",
+        value=clf_logistic_model.n_features_in_,
+        delta=None,
+        delta_color="normal",
+    )
+
+    st.subheader("Logistic Regression Coefficient Values")
+    st.write(split_dataset)
+    st.write(type(split_dataset))
+
+    coef_dict = create_coeff_dict_logistic_model(
+        clf_logistic_model, split_dataset.X_y_train)
+
+    df = coeff_dict_to_sorted_df(coef_dict)
+
+    fig = plot_logistic_coeff_barh(df)
+
+    st.plotly_chart(fig)
+
+
+def logistic_train_model(split_dataset: SplitDataset):
+    st.header("Logistic Regression Model")
+
+    clf_logistic_model = create_clf_logistic_model(
+        split_dataset.X_train, split_dataset.y_train
+    )
+
+    interpret_clf_logistic_model(clf_logistic_model, split_dataset)
+
+    return clf_logistic_model

src/visualization/graphs_logistic.py

@@ -0,0 +1,12 @@
+import plotly.express as px
+
+
+def plot_logistic_coeff_barh(df):
+    fig = px.bar(data_frame=df, x="Value",
+                 y="Coefficient", orientation="h")
+
+    fig.update_layout(
+        title="Logistic Regression Coefficients",
+        xaxis_title="Value",
+        yaxis_title="Coefficient",)
+    return fig