离散:
def max_count(df: pd.DataFrame):
return df.value_counts().index[0]
look_cols = ['col_1', 'col_2', 'col_3',]
for col in look_cols:
print("| ", col, " |")
print(data.groupby('type')[col].apply(max_count))
print('---------------------')
连续:
偏态特征: kurt, kurtosis, mad(mean absolute deviation),
大小值,分位数:min, max, median,
统计特征:mean, mode, std, var,
最大最小值:nlargest, nsmallest, idxmax, idxmin
数学运算:sum
# 最大
print(data.groupby('type')[your_col].max())
# 均值
print(data.groupby('type')[col].mean())
参考:
https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.mad.html