autoML.py

import streamlit as st
import pandas as pd
import numpy as np
from flaml import AutoML
from flaml.automl.data import get_output_from_log
import pickle
import plotly.express as px
import base64
import time

from utils import csv_to_featuers_list, pre_process_df, pre_process_features

def autoML(csv, task, budget, label, metric_to_minimize_class, metric_to_minimize_reg):

    progress_text="Training in progress. Please wait."
    my_bar = st.progress(0, text=progress_text)
    time.sleep(0.5)

    df = pd.read_csv(csv)
    df = pre_process_df(df)
    df_features = df[df.columns.difference([label])]
    df_features=(df_features-df_features.mean())/df_features.std()
    y = df[label]

    my_bar.progress(50, text=progress_text)


    if task == 'Classification':
        automl_settings = {
            "time_budget": int(budget),
            "metric": metric_to_minimize_class,
            "task": 'classification',
            "log_file_name": 'classlog.log',
            "early_stop": True,
            "eval_method": "holdout"
        }

    if task == 'Regression':
        automl_settings = {
            "time_budget": int(budget),
            "metric": metric_to_minimize_reg,
            "task": 'regression',
            "log_file_name": 'reglog.log',
            "early_stop": True,
            "eval_method": "holdout"
        }

    automl = AutoML()
    automl.fit(df_features, y, **automl_settings)

    my_bar.progress(100, text=progress_text)
    time.sleep(0.5)
    my_bar.empty()

    tab1, tab2 = st.tabs(["AutoML", "Best Model"])

    with tab1:
        
        if task == 'Classification': 
            log = 'classlog.log'
            metric = metric_to_minimize_class
        if task == 'Regression': 
            log = 'reglog.log'
            metric = metric_to_minimize_reg

        time_history, best_valid_loss_history, valid_loss_history, config_history, metric_history = get_output_from_log(filename=log, time_budget=120)
    
        def model(s):
            mod = s.get('Current Learner')
            return mod
        
        def hp(s):
            hparams = s.get('Current Hyper-parameters')
            return hparams

        df_res = pd.DataFrame({'time': time_history, 
                            metric: 1 - np.array(best_valid_loss_history),
                            'model': list(map(model, config_history)),
                            })
        
        fig = px.line(df_res,
                    title='evolution of best models found by AutoML',
                    x='time',
                    y=metric,
                    hover_name='model',
                    line_shape='hv',
                    range_y=[0,1])
        
        st.plotly_chart(fig, theme="streamlit")

        models = pd.DataFrame({'learner': list(map(model, config_history))})
        hps = list(map(hp, config_history))
        df_hp = pd.DataFrame(hps)
        df_models = pd.concat((models, df_hp), axis=1)

        def highlight_last_row(s):
            return ['background-color: yellow' if i == len(s) - 1 else '' for i in range(len(s))]

        st.dataframe(df_models.style.apply(highlight_last_row, axis=0))

        st.write('Estimator tested')
        st.table(automl.estimator_list)

    with tab2:
        st.header('Best Model')

        st.text(automl.model.estimator)

        col1, col2, col3 = st.columns((1,1,1))

        with col1:
            st.metric(label="r2_score", value=round(1 - automl.best_loss, 2))
        with col2:
            st.metric(label="Time to find", value=str(round(automl.time_to_find_best_model, 2))+' sec')
        with col3:
            st.metric(label="Time to train", value=str(round(automl.best_config_train_time, 2))+' sec')

        df_features_importance = pd.DataFrame({'features name': automl.model.estimator.feature_name_, 'features importance': automl.model.estimator.feature_importances_})
        fig_features = px.bar(df_features_importance, x='features importance', y='features name')

        st.divider()
        st.plotly_chart(fig_features, theme="streamlit")


        def download_model(model):
            output_model = pickle.dumps(model)
            b64 = base64.b64encode(output_model).decode()
            href = f'<a href="data:file/output_model;base64,{b64}" download="automl.pkl">Download Trained Model File (.pkl)</a>'
            st.markdown(href, unsafe_allow_html=True)

        download_model(automl)