src/eda.py

# eda.py
import streamlit as st
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.impute import SimpleImputer

def run_eda():
    st.header("📊 Exploratory Data Analysis")

    df_raw = st.session_state.original_df
    df = st.session_state.processed_df

    c1, c2 = st.columns(2)
    with c1:
        st.subheader("Raw Data")
        st.dataframe(df_raw.head(10))
        st.write(df_raw.describe(include="all"))

    with c2:
        st.subheader("Processed Data")
        st.dataframe(df.head(10))

    st.subheader("🎯 Feature & Target Selection")

    cols = df.columns.tolist()
    target = st.selectbox("Target", cols)
    features = st.multiselect("Features", [c for c in cols if c != target])

    st.session_state.target_col = target
    st.session_state.feature_cols = features

    st.subheader("🧹 Cleaning")

    if st.checkbox("Apply smart imputation"):
        num = df.select_dtypes(include=np.number).columns
        cat = df.select_dtypes(exclude=np.number).columns

        if len(num):
            df[num] = SimpleImputer(strategy="mean").fit_transform(df[num])
        if len(cat):
            df[cat] = SimpleImputer(strategy="most_frequent").fit_transform(df[cat])

        st.session_state.processed_df = df
        st.success("Imputation complete")
        st.rerun()

    st.subheader("📈 Visuals")

    plot = st.selectbox(
        "Plot type",
        ["Correlation Heatmap", "Target Distribution"]
    )

    fig, ax = plt.subplots(figsize=(8,6))

    if plot == "Correlation Heatmap":
        sns.heatmap(df.select_dtypes(np.number).corr(), annot=True, ax=ax)

    elif plot == "Target Distribution":
        sns.histplot(df[target], kde=True, ax=ax)

    st.pyplot(fig)
    plt.close(fig)