Chaining Streamlit forms and passing values using session state API

multiform_chaining.py

import streamlit as st
from streamlit_pandas_profiling import st_profile_report
from streamlit import session_state as session

import numpy as np
import pandas as pd
from pandas_profiling import ProfileReport

if 'data_uploader_submitted' not in session:
    session.data_uploader_submitted = False
if 'cluster_duplicates_submitted' not in session:
    session.cluster_duplicates_submitted = False

def cluster_duplicates(df, col_name, dis_num, dis_non_alphanum, sim, aff):
    st.write(col_name)
    st.write(df.head())
    st.write(df[col_name].unique())

@st.experimental_memo(show_spinner=True, persist='disk')
def get_profile_report(file_info, df):
    pr = ProfileReport(df, explorative=True, lazy=True, minimal=True)
    return pr

def profiler(file, delim):
    file = st.session_state.upload
    delimiter = st.session_state.delim.split(" ")[1][1:-1]
    df = pd.read_csv(file, sep=delimiter, engine="python")
    session['df'] = df
    file_info = {"Filename": file.name, "FileType": file.type, "FileSize": file.size}
    st.write(file_info)
    pr = get_profile_report(file_info, df)
    st_profile_report(pr)

    cluster_duplicates_form = st.form(key="cluster_duplicates")
    with cluster_duplicates_form:
        cols = [val for val in df.columns]
        col_name = st.selectbox("Select column for clustering", cols, key="col_name")
        dis_num = st.checkbox("discard_numeric", key="dis_num")
        dis_non_alphanum = st.checkbox("discard_nonalpha_numeric", key="dis_non_alphanum")
        similarity = st.radio(label="Select Similarity Measure",
                              options=["levenshtein (recommended)", "cosine", "jaro_winkler", "trigram",
                                       "levenshtein_partial"], key="similarity")
        affinity = st.radio(label="Select Distance Measure",
                       options=["euclidean", "l1", "l2", "manhattan", "cosine", "precomputed"], key="affinity")
        if cluster_duplicates_form.form_submit_button(label = "Cluster Duplicates"):
            session.cluster_duplicates_submitted = True

def data_uploader_form():
    file_upload_form = st.form(key="file_upload")
    with file_upload_form:
        data_file = st.file_uploader("Upload File", type=['csv', 'xlsx'], key="upload")
        delim_list = ["pipe (|)", r"tab (\t)", "comma (,)", "semicolon (;)"]
        delim = st.selectbox("Select File Seperator/Delimiter", delim_list, key="delim")
        if file_upload_form.form_submit_button(label='Profile Data'):
            session.data_uploader_submitted = True

if __name__ =="__main__":
    #st.set_page_config(layout="wide")
    st.write("Data Profiler :wave:")
    data_uploader_form()
    if session.data_uploader_submitted:
        profiler(session.upload, session.delim)
    if session.cluster_duplicates_submitted:
        method_args = (session.df, session.col_name, session.dis_num, session.dis_non_alphanum, session.similarity, session.affinity)
        print(method_args)
        print(list(session.items()))
        cluster_duplicates(*method_args)