Function for Normalizing pandas DataFrame

scale.py

def normalize(df):
  """
  Function for min-max Scaling a pandas DataFrame
  @param:
  Takes a pandas DataFrame: df
  Returns: a normalized DataFrame
  along with a dict containing rescaling 
  coef which can be used in below function.
  """
    result = df.copy()
    param = dict()
    for feature_name in df.columns:
        param['max_value_'+str(feature_name)] = df[feature_name].max()
        param['min_value_'+str(feature_name)] = df[feature_name].min()
        max_value = df[feature_name].max()
        min_value = df[feature_name].min()
        result[feature_name] = (df[feature_name] - min_value) / (max_value - min_value)
    return result, param

def normalize_zscore(df):
  """
  Function for Z-score Scaling a pandas DataFrame
  @param:
  Takes a pandas DataFrame: df
  Returns: a normalized DataFrame
  along with a dict containing rescaling 
  coef which can be used in below function.
  """
    result = df.copy()
    param = dict()
    for feature_name in df.columns:
        param['mean_'+str(feature_name)] = df[feature_name].mean()
        param['std_'+str(feature_name)] = df[feature_name].std()
        mean = df[feature_name].mean()
        std = df[feature_name].std()
        result[feature_name] = (df[feature_name] - mean) / std
    return result, param

def denorm_zscore(normalizedDF, param):
  """
  Function for rescaling a scaled DataFrame
  @param:
  Takes a pandas DataFrame and a dict output from above two functions
  Returns:
  Returns a rescaled DataFrame
  """
    for feature_name in normalizedDF.columns:
        mean = param["mean_"+str(feature_name)]
        std = param["std_"+str(feature_name)]
        normalizedDF[feature_name] = (normalizedDF[feature_name] * std) + mean
    return normalizedDF