app.py

import gradio as gr
from datasets import load_dataset, Dataset
import pandas as pd
from huggingface_hub import create_repo
from huggingface_hub import login
login(token='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')

dataset = load_dataset("torileatherman/sentiment_analysis_batch_predictions", split='train')
predictions_df = pd.DataFrame(dataset)
grouped_predictions = predictions_df.groupby(predictions_df.Prediction)
positive_preds = grouped_predictions.get_group(2)
neutral_preds = grouped_predictions.get_group(1)
negative_preds = grouped_predictions.get_group(0)

predictions_df['Prediction'] = predictions_df['Prediction'].map({0: 'Negative', 1: 'Neutral', 2: 'Positive'})


def article_selection(sentiment):
    if sentiment == "Positive":
        predictions = positive_preds
        top3 = predictions[0:3]
        top3_result = top3[['Headline_string','Url']]
        top3_result.rename(columns = {'Headline_string':'Headlines', 'Url':'URL'})
        return top3_result
    
    elif sentiment == "Negative":
        predictions = negative_preds
        top3 = predictions[0:3]
        top3_result = top3[['Headline_string','Url']]
        top3_result.rename(columns = {'Headline_string':'Headlines', 'Url':'URL'})
        return top3_result
    else:
        predictions = neutral_preds
        top3 = predictions[0:3]
        top3_result = top3[['Headline_string','Url']]
        top3_result.rename(columns = {'Headline_string':'Headlines', 'Url':'URL'})
        return top3_result

def manual_label():
    # Selecting random row from batch data
    random_sample = predictions_df.sample()
    #random_sample.to_csv('/Users/torileatherman/Github/ID2223_scalable_machine_learning/news_articles_sentiment/sample.csv', index=False)
    random_sample.to_csv('torileatherman/sample')
    random_headline = random_sample['Headline_string'].iloc[0]
    random_prediction = random_sample['Prediction'].iloc[0]
    return random_headline, random_prediction


def thanks(sentiment):
    labeled_sentiments = []
    labeled_sentiments.append(sentiment)
    #counter = len(labeled_sentiments)
    #counter = str(counter)
    #login(token = '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')
    #create_repo("torileatherman/"+counter+"labeled_data")
    labeled_sentiments = pd.DataFrame(labeled_sentiments, columns = ['Manual Predictions'])
    labeled_sentiments.to_csv('/Users/torileatherman/Github/ID2223_scalable_machine_learning/news_articles_sentiment/manual_labels.csv', index=False)
    #labeled_sentiments = Dataset.from_pandas(labeled_sentiments)
    #labeled_sentiments.push_to_hub("torileatherman/"+counter+"labeled_data")
    return f"""Thank you for making our model better!"""
    

description1 =  '''
        This application recommends news articles depending on the sentiment of the headline.
        Enter your preference of what type of news articles you would like recommended to you today: Positive, Negative, or Neutral.
        '''

description2 =  '''
        This application will show you a random news headline and our predicted sentiment.
        In order to improve our model, mark the real sentiment of this headline!
        '''

suggestion_demo = gr.Interface(
    fn=article_selection,
    title = 'Recommending News Articles',
    inputs = gr.Dropdown(["Positive","Negative","Neutral"], label="What type of news articles would you like recommended?"),
    outputs = "dataframe",
    #outputs = [gr.Textbox(label="Recommended News Articles (1/3)"),gr.Textbox(label="Recommended News Articles (2/3)"),gr.Textbox(label="Recommended News Articles (3/3)")],
    description = description1
)

with gr.Blocks() as manual_label_demo:
    description = description2
    generate_btn = gr.Button('Show me a headline!')
    generate_btn.click(fn=manual_label, outputs=[gr.Textbox(label="News Headline"),gr.Textbox(label="Our Predicted Sentiment")])
    drop_down_label = gr.Dropdown(["Positive","Negative","Neutral"], label="Select the true sentiment of the news article.")
    submit_btn =  gr.Button('Submit your sentiment!')
    submit_btn.click(fn=thanks, inputs=drop_down_label, outputs=gr.Textbox())

manual_label_demo1 = gr.Interface(
    fn=thanks,
    title="Manually Label a News Article",
    inputs=[gr.Textbox(label = "Paste in URL of news article here."),
            gr.Dropdown(["Positive","Negative","Neutral"], label="Select the sentiment of the news article.")],
    outputs = gr.Textbox(label="Output"),
    description = description2
)


demo = gr.TabbedInterface([suggestion_demo, manual_label_demo], ["Get recommended news articles", "Help improve our model"])


demo.launch()