Comprehensive List of Aggregation Functions and Custom Feature Engineering Utilities for Python

Recently readers asked about effective feature engineering techniques for beginners, especially aggregation operations that can boost early competition performance.

The article lists built‑in pandas aggregation functions such as mean() , sum() , size() , count() , std() , var() , sem() , first() , last() , nth() , min() , and max() , and mentions other important aggregation utilities.

It also provides a collection of custom Python functions for time‑series feature extraction, including statistical measures (median, variation_coefficient, variance, skewness, kurtosis, standard_deviation, large_standard_deviation), drawdown and drawup calculations, peak detection, count and ratio metrics, and many helper utilities such as count_above , count_below , number_peaks , mean_abs_change , root_mean_square , abs_energy , and others.

All functions are presented in raw code format, preserving their definitions for direct reuse:

def median(x):
    return np.median(x)

def variation_coefficient(x):
    mean = np.mean(x)
    if mean != 0:
        return np.std(x) / mean
    else:
        return np.nan

# ... (additional function definitions omitted for brevity) ...

Additionally, the article shows how to group these functions into logical categories for feature generation, e.g.,

base_stats = ['mean','sum','size','count','std','first','last','min','max',median,skewness,kurtosis]
higher_order_stats = [abs_energy,root_mean_square,sum_values,realized_volatility,realized_abs_skew,realized_skew,realized_vol_skew,realized_quarticity]
# ... (other groupings) ...

Reference: https://www.kaggle.com/code/lucasmorin/amex-feature-engineering-2-aggreg-functions