app.py

import gradio as gr
import numpy as np
import sklearn.datasets as d
from sklearn.linear_model import *
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import *
from sklearn.utils import all_estimators


import inspect
import pandas as pd
import sklearn.metrics as m

def predict(dataset, model, split, metrics):
  pass
  models = [cls for cls in all_estimators() if cls[0] == model]
  if len(models) == 0:
    return "Model not found"
  model = models[0][1]()
  data = getattr(d, dataset)()
  X_train, X_test, y_train, y_test = train_test_split(data.data, data.target, test_size=float(split))
  scaler = StandardScaler()
  X_train = scaler.fit_transform(X_train)
  X_test = scaler.transform(X_test)
  model.fit(X_train, y_train)
  pred = model.predict(X_test)
  df = pd.DataFrame(data.data, columns=data.feature_names)
  df["target"] = data.target
  scores = []
  for metric in metrics:
    try:
      if hasattr(m, metric):
        scores.append((metric, getattr(m, metric)(y_test, pred)))
    except:
      pass
  scoress = pd.DataFrame(scores, columns=["metric", "score"])
  return gr.Dataframe(scoress, headers=scoress.columns.tolist(), datatype=["numeric"] * len(df.columns))

demo = gr.Interface(fn=predict, inputs=[
    gr.Dropdown([ name for name, obj in inspect.getmembers(d)
    if inspect.isfunction(obj) and not name.startswith("_")], value="load_breast_cancer", label="Dataset"), 
    gr.Dropdown([name for name, cls in all_estimators()], value="RandomForestClassifier", label="Model"), 
    gr.Textbox(value="0.2", label="TrainTest Split"),
    gr.CheckboxGroup([n for n in dir(m) if callable(getattr(m, n)) and not n.startswith("_")], label="metrics", value="accuracy_score")
], outputs="dataframe")

demo.launch(share=True, debug=True)