import gradio as gr
import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.cluster import KMeans

labels = {}

def get_sentiment(input_txt, k):
  data = input_txt.splitlines()
  df = pd.DataFrame(data, columns=["content"])
  vec = TfidfVectorizer()
  features = vec.fit_transform(df["content"])
  kmeans = KMeans(k, n_init=10, random_state=42)
  kmeans.fit(features)
  df['kmeans'] = kmeans.labels_
  labels = df['kmeans'].value_counts(normalize=True).to_dict()
  return labels

iface = gr.Interface(fn= get_sentiment,  
                     inputs = [gr.inputs.Textbox(lines=8, placeholder="News Here..."), gr.inputs.Slider(1, 101)],
                     outputs = ["label"], 
                     title = "Short Document Clustering")

if __name__ == "__main__":
    app, local_url, share_url = iface.launch()