o

README.md

@@ -1,4 +1,4 @@
-# (UPDATE IN README.md) Project Title
+# (UPDATE IN README.md) Saas Growth prediction
 
 (UPDATE IN README.md)
 Briefly describe the purpose/result(s) of your project, the skills you applied, and the AI4ALL Ignite program.

fair_model.py

@@ -0,0 +1,222 @@
+import os
+import joblib
+import pandas as pd
+import streamlit as st
+
+import matplotlib.pyplot as plt
+from sklearn.model_selection import train_test_split
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.metrics import (
+    accuracy_score,
+    classification_report,
+    roc_auc_score,
+    confusion_matrix,
+    roc_curve,
+)
+
+# -------------------------
+# APP LAYOUT
+# -------------------------
+
+st.title("Fair Credit Card Default Prediction (No Demographics)")
+st.write(
+    "This app trains a Random Forest classifier on the credit card default dataset, "
+    "removing demographic features (SEX, EDUCATION, MARRIAGE, AGE) to improve fairness."
+)
+
+# -------------------------
+# DATA LOADING
+# -------------------------
+
+st.sidebar.header("Data Settings")
+
+default_filename = "Credit Card Defaulter Prediction.csv"
+use_uploaded = st.sidebar.checkbox("Upload CSV instead of using local file")
+
+if use_uploaded:
+    uploaded_file = st.sidebar.file_uploader("Upload CSV file", type=["csv"])
+    if uploaded_file is not None:
+        df = pd.read_csv(uploaded_file)
+    else:
+        st.stop()
+else:
+    if not os.path.exists(default_filename):
+        st.error(
+            f"File `{default_filename}` not found. "
+            f"Place it in the same folder or upload a CSV using the sidebar."
+        )
+        st.stop()
+    df = pd.read_csv(default_filename)
+
+st.subheader("Raw Data (first 5 rows)")
+st.dataframe(df.head())
+
+st.write("**Data Summary**")
+st.write(df.describe(include="all"))
+
+# -------------------------
+# CLEANING / TARGET
+# -------------------------
+
+# Drop ID so that the model does not memorize IDs
+if "ID" in df.columns:
+    df = df.drop(["ID"], axis=1)
+
+target = "default "  # note the space to match your original script
+if target not in df.columns:
+    st.error(f"Target column `{target}` not found in data.")
+    st.stop()
+
+# Change Y and N to 1 and 0
+df[target] = df[target].map({"N": 0, "Y": 1})
+
+st.write("**Target value counts (after mapping N→0, Y→1):**")
+st.bar_chart(df[target].value_counts())
+
+# -------------------------
+# FAIRNESS: DROP DEMOGRAPHICS
+# -------------------------
+
+sens_cols = ["SEX", "EDUCATION", "MARRIAGE", "AGE"]
+present_sens_cols = [c for c in sens_cols if c in df.columns]
+
+st.subheader("Fairness Setup")
+st.write(
+    "We **remove** the following sensitive / demographic columns before training "
+    "to reduce direct dependence on them:"
+)
+st.write(present_sens_cols if present_sens_cols else "None of the listed sensitive columns are in the dataset.")
+
+df_fair = df.drop(columns=present_sens_cols, errors="ignore")
+
+X_fair = df_fair.drop(columns=[target])
+y_fair = df[target]
+
+# -------------------------
+# TRAIN / TEST SPLIT & MODEL
+# -------------------------
+
+test_size = st.sidebar.slider("Test size (fraction)", 0.1, 0.4, 0.2, 0.05)
+random_state = st.sidebar.number_input("Random state", value=42, step=1)
+
+X_train_f, X_test_f, y_train_f, y_test_f = train_test_split(
+    X_fair,
+    y_fair,
+    test_size=test_size,
+    random_state=random_state,
+    stratify=y_fair,
+)
+
+st.write("**Feature count (after removing sensitive columns):**", X_fair.shape[1])
+
+n_estimators = st.sidebar.slider("Number of trees (n_estimators)", 50, 300, 100, 10)
+
+rf_fair = RandomForestClassifier(
+    n_estimators=n_estimators,
+    class_weight="balanced",
+    random_state=random_state,
+    n_jobs=-1,
+)
+
+if st.button("Train Fair Model"):
+    rf_fair.fit(X_train_f, y_train_f)
+
+    y_pred_f = rf_fair.predict(X_test_f)
+    y_proba_f = rf_fair.predict_proba(X_test_f)[:, 1]
+
+    # Save model
+    joblib.dump(rf_fair, "fair_rf_model.sav")
+
+    st.success("Model trained and saved as `fair_rf_model.sav`.")
+
+    # -------------------------
+    # METRICS
+    # -------------------------
+
+    st.subheader("Metrics")
+
+    acc = accuracy_score(y_test_f, y_pred_f)
+    roc = roc_auc_score(y_test_f, y_proba_f)
+    cm = confusion_matrix(y_test_f, y_pred_f)
+
+    st.write(f"**Accuracy:** {acc:.4f}")
+    st.write(f"**ROC-AUC:** {roc:.4f}")
+
+    st.write("**Classification Report:**")
+    st.text(classification_report(y_test_f, y_pred_f))
+
+    # -------------------------
+    # VISUAL 1: CONFUSION MATRIX
+    # -------------------------
+
+    st.subheader("Confusion Matrix")
+
+    fig_cm, ax_cm = plt.subplots()
+    ax_cm.imshow(cm, interpolation="nearest")
+    ax_cm.set_title("Confusion Matrix")
+    ax_cm.set_xlabel("Predicted label")
+    ax_cm.set_ylabel("True label")
+
+    tick_marks = [0, 1]
+    ax_cm.set_xticks(tick_marks)
+    ax_cm.set_yticks(tick_marks)
+    ax_cm.set_xticklabels(["No Default (0)", "Default (1)"])
+    ax_cm.set_yticklabels(["No Default (0)", "Default (1)"])
+
+    # Label each cell
+    for i in range(cm.shape[0]):
+        for j in range(cm.shape[1]):
+            ax_cm.text(
+                j,
+                i,
+                cm[i, j],
+                ha="center",
+                va="center",
+            )
+
+    st.pyplot(fig_cm)
+
+    # -------------------------
+    # VISUAL 2: ROC CURVE
+    # -------------------------
+
+    st.subheader("ROC Curve")
+
+    fpr, tpr, thresholds = roc_curve(y_test_f, y_proba_f)
+
+    fig_roc, ax_roc = plt.subplots()
+    ax_roc.plot(fpr, tpr, label=f"ROC curve (AUC = {roc:.2f})")
+    ax_roc.plot([0, 1], [0, 1], linestyle="--", label="Random")
+    ax_roc.set_xlabel("False Positive Rate")
+    ax_roc.set_ylabel("True Positive Rate")
+    ax_roc.set_title("ROC Curve")
+    ax_roc.legend(loc="lower right")
+
+    st.pyplot(fig_roc)
+
+    # -------------------------
+    # VISUAL 3: FEATURE IMPORTANCE
+    # -------------------------
+
+    st.subheader("Feature Importances (Fair Model)")
+
+    importances = rf_fair.feature_importances_
+    feat_names = X_fair.columns
+
+    # Sort by importance
+    sorted_idx = importances.argsort()[::-1]
+    sorted_importances = importances[sorted_idx]
+    sorted_names = feat_names[sorted_idx]
+
+    fig_imp, ax_imp = plt.subplots(figsize=(8, 5))
+    ax_imp.bar(range(len(sorted_importances)), sorted_importances)
+    ax_imp.set_xticks(range(len(sorted_importances)))
+    ax_imp.set_xticklabels(sorted_names, rotation=90)
+    ax_imp.set_ylabel("Importance")
+    ax_imp.set_title("Random Forest Feature Importances")
+
+    plt.tight_layout()
+    st.pyplot(fig_imp)
+
+else:
+    st.info("Click **Train Fair Model** to fit the model and see graphs.")