import gradio as gr
import pandas as pd
from sklearn import datasets
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.preprocessing import LabelEncoder

def findCorrelation(dataset, target):
    
  df = pd.read_csv(dataset.name)
  
  non_numeric_cols = df.select_dtypes('object').columns.tolist()
  
  for non_numeric_col in non_numeric_cols:
    label_encoder = LabelEncoder()
    df[non_numeric_col] = label_encoder.fit_transform(df[non_numeric_col])
  
  d = df.corr()[target].to_dict()
  d.pop(target)
  
  keys = sorted(d.items(), key=lambda x: x[0], reverse=True) 
    
  fig1 = plt.figure()
  hm = sns.heatmap(df.corr(), annot = True)
  hm.set(title = "Correlation matrix of dataset\n")
    
  fig2 = plt.figure()
  sns.regplot(x=df[keys[0][0]], y=df[target])
  
  fig3 = plt.figure()
  sns.regplot(x=df[keys[1][0]], y=df[target])

  fig4 = plt.figure()
  sns.regplot(x=df[keys[2][0]], y=df[target])
   
  return d, fig1, fig2, fig3, fig4

demo = gr.Interface(fn=findCorrelation, inputs=[gr.File(), 'text'], outputs=[gr.Label(num_top_classes = 10), gr.Plot(), gr.Plot(), gr.Plot(), gr.Plot()], title="Find correlation")
demo.launch(debug=True)