HarshulNanda/HARM_ML_App_ludwig

1import os2from pytube import YouTube3import pytube4from stqdm import stqdm5import pandas as pd6from youtubesearchpython import Video, ResultMode7import streamlit as st8import scrapetube9from categoryPredictor import predictCategoryFor10import pandas as pd11 12@st.experimental_memo13def convert_df(df):14   return df.to_csv(index=False).encode('utf-8')15 16def generate_channel_video_data(of_channel, with_number_of_videos):17    video_urls = []18    c_id = Video.get(of_channel, mode=ResultMode.json, get_upload_date=True)["channel"]["id"]19    videos = scrapetube.get_channel(c_id)20    i = 021    for video in videos:22        video_urls.append("https://www.youtube.com/watch?v="+str(video['videoId']))23        i += 124        if i == with_number_of_videos:25            break26 27    data = {28        "Title": [],29        "Description": [],30        "Category": [],31        "Is Educational?": [],32        "Beyond Exams Category": [],33    }34 35    timer = stqdm(video_urls)36 37    for video in timer:38        timer.set_description("☕️ Have a coffee, while we are generating your dataset.  ")39        try:40            v = Video.get(video, mode = ResultMode.json, get_upload_date=True)41            t = v["title"]42            d = v["description"]43            c = v["category"]44            isEdu, isCat, cat_array, sub_array = predictCategoryFor(video)45            data["Description"].append(d)46            data["Category"].append(c)47            data["Title"].append(t)48            data["Is Educational?"].append(isEdu)49            data["Beyond Exams Category"].append(isCat)50        except Exception as e:51            print(e)52            continue53 54    df = pd.DataFrame(data)55    st.dataframe(df)56    csv = convert_df(df)57 58    st.download_button(59        "Download this dataframe",60        csv,61        "file.csv",62        "text/csv",63        key='download-csv'64    )