Extreme outlier detection

outlier_function.py

def outlier_function(df, col_name):
    ''' this function detects first and third quartile and interquartile range for a given column of a dataframe
    then calculates upper and lower limits to determine outliers conservatively
    returns the number of lower and uper limit and number of outliers respectively
    '''
    first_quartile = np.percentile(np.array(df[col_name].tolist()), 25)
    third_quartile = np.percentile(np.array(df[col_name].tolist()), 75)
    IQR = third_quartile - first_quartile
                      
    upper_limit = third_quartile+(3*IQR)
    lower_limit = first_quartile-(3*IQR)
    outlier_count = 0
                      
    for value in df[col_name].tolist():
        if (value < lower_limit) | (value > upper_limit):
            outlier_count +=1
    return lower_limit, upper_limit, outlier_count