import os
import streamlit as st
import glob
import pandas as pd
import matplotlib.pyplot as plt

def run_eda():
    st.title('CNN Image Classifier')
    
    # Set directories
    train_dir = "Data/train"

    

    # train sets
    for files in os.listdir(train_dir):
        print(os.path.join(train_dir,files))

    num_A = len(glob.glob(os.path.join(train_dir, "normal", "*.png")))
    num_ad = len(glob.glob(os.path.join(train_dir, "adenocarcinoma_left.lower.lobe_T2_N0_M0_Ib", "*.png")))
    num_sq = len(glob.glob(os.path.join(train_dir, "squamous.cell.carcinoma_left.hilum_T1_N2_M0_IIIa", "*.png")))
    num_lg = len(glob.glob(os.path.join(train_dir, "large.cell.carcinoma_left.hilum_T2_N2_M0_IIIa", "*.png")))


    class_names = ['Normal', 'Adenocarcinoma', 'Squamous', 'Large Cell']
    num_images = [num_A, num_ad, num_sq, num_lg]

    fig, ax = plt.subplots()
    ax.bar(class_names, num_images)
    ax.set_title('Banyaknya Data Setiap Kelas pada Dataset Train')
    ax.set_xlabel('Kelas')
    ax.set_ylabel('Jumlah Data')
    st.pyplot(fig)