import streamlit as st
import pandas as pd
import io
import base64
from sklearn.impute import SimpleImputer


st.set_page_config(page_title="CSV Data Cleaning Tool")

hide_streamlit_style = """
            <style>
            #MainMenu {visibility: hidden;}
            footer {visibility: hidden;}
            </style>
            """
st.markdown(hide_streamlit_style, unsafe_allow_html=True)

st.title("CSV Data Tool")

st.markdown("กดเลือกหัว Tool ข้อที่ต้องการจะใช้ได้เลยนะจ๊ะ")

uploaded_files = st.file_uploader("Choose CSV files", type="csv", accept_multiple_files=True)

dataframes = []

if uploaded_files:
    for file in uploaded_files:
        file.seek(0)
        df = pd.read_csv(file)
        dataframes.append(df)

    st.markdown("Data Cleansing")

    duplicate_columns = st.checkbox("Remove duplicate columns", value=False)
    if duplicate_columns :
        for i, df in enumerate(dataframes):
            dataframes[i] = df.drop_duplicates(inplace=False)
    
    remove_empty_rows = st.checkbox("Remove empty rows", value=False)
    if remove_empty_rows:
        for i, df in enumerate(dataframes):
            dataframes[i] = df.dropna(how="all", inplace=False)
    
    impute_mean = st.checkbox("Impute missing values with mean (for int and float columns)",value=False)
    if impute_mean:
        for i, df in enumerate(dataframes):
            numeric_cols = df.select_dtypes(include=['int', 'float']).columns
            imputer = SimpleImputer(strategy='mean')
            df[numeric_cols] = imputer.fit_transform(df[numeric_cols])
            dataframes[i] = df                       

    impute_most_frequent = st.checkbox("Impute missing values with most frequent category (for categorical columns)",value=False)
    if impute_most_frequent:
        for i, df in enumerate(dataframes):
            categorical_cols = df.select_dtypes(include=['object']).columns
            imputer = SimpleImputer(strategy='most_frequent')
            df[categorical_cols] = imputer.fit_transform(df[categorical_cols])
            dataframes[i] = df

    
    selected_out = st.selectbox("เลือก columns ที่จะดู Outlier", df.columns)
    if selected_out:
        col = selected_out
        st.write(f"คอลัมน์ {col}:")
        # Calculate Z-Scores for the selected column
        z_scores = np.abs((df[col] - df[col].mean()) / df[col].std())
        # Set a threshold for identifying outliers (e.g., z_score > 3)
        threshold = 3
        # Identify outliers
        outliers = df[z_scores > threshold]
        st.write("Outliers:")
        st.write(outliers)

    
    st.markdown("Data transform")
    
    for i, df in enumerate(dataframes):
        st.dataframe(df)
    
    selected_values = st.multiselect("เลือกค่าจากคอลัมน์", df.columns)

    
    convert_to_String = st.checkbox("convert columns to String", value=False)
    convert_to_float = st.checkbox("convert columns to Float", value=False)
    if convert_to_String:
        df[selected_values] = df[selected_values].astype(str)
    if convert_to_float:
        df[selected_values] = df[selected_values].astype(float)

    
    

    
    show_dataframes = st.checkbox("Show DataFrames", value=True)
    if show_dataframes:
        for i, df in enumerate(dataframes):
            st.write(f"DataFrame {i + 1}")
            st.dataframe(df)
            

    
if st.button("Download cleaned data"):
    for i, df in enumerate(dataframes):
        csv = df.to_csv(index=False)
        b64 = base64.b64encode(csv.encode()).decode()
        href = f'<a href="data:file/csv;base64,{b64}" download="cleaned_data_{i + 1}.csv">Download cleaned_data_{i + 1}.csv</a>'
        st.markdown(href, unsafe_allow_html=True)


st.markdown("")
st.markdown("---")
st.markdown("")