Handle ML functions

handy_ml_functions.py

# Usage:
# !git clone https://gist.github.com/d828518b070af514c339d9efa548f525.git gist_code
# !ls -la gist_code
# import glob, importlib.util, sys
# 
# helper_path = glob.glob("gist_code/*.py")[0]
# spec = importlib.util.spec_from_file_location("helpers", helper_path)
# helpers = importlib.util.module_from_spec(spec)
# sys.modules["helpers"] = helpers
# spec.loader.exec_module(helpers)
# 
# print("Loaded:", helper_path)
# print("Found functions:", [n for n in dir(helpers) if not n.startswith("_")])

def plot_label_distribution(df, label_col):
  import seaborn as sns, matplotlib.pyplot as plt
  plt.figure(figsize=(8, 4))
  sns.countplot(y=df[label_col], order=df[label_col].value_counts().index)
  plt.title("Label distribution")
  plt.tight_layout()
  plt.show()
  
 # kaggle_dataset_download("a_name", file_type="*.csv")
def kaggle_dataset_download(dataset_name: str, file_type: str):
  import kagglehub
  import os, glob
  path = kagglehub.dataset_download(dataset_name)
  csv_files = glob.glob(os.path.join(path, file_type))
  return csv_files