变量有重复值怎么分箱？

基于分位数分箱

     import pandas as pd
     import numpy as np

     data = pd.Series(np.random.randint(1, 100, size=100))
     data_with_duplicates = pd.concat([data, data[:20]])  # 添加一些重复值

     # 去除重复值
     data_unique = data_with_duplicates.drop_duplicates()

     quantiles = data_unique.quantile([0.25, 0.5, 0.75])
     bins = [data_unique.min(), quantiles[0.25], quantiles[0.5], quantiles[0.75], data_unique.max()]
     labels = ['Bin_1', 'Bin_2', 'Bin_3', 'Bin_4']
     data_binned_unique = pd.cut(data_unique, bins=bins, labels=labels)

     print(f"原始有重复值数据样本数量：{len(data_with_duplicates)}")
     print(f"去除重复值后分箱数据样本数量：{len(data_binned_unique)}")

等宽分箱

     min_value = data_with_duplicates.min()
     max_value = data_with_duplicates.max()
     bin_width = (max_value - min_value) / 4
     bins = [min_value + i * bin_width for i in range(5)]
     labels = ['Bin_1', 'Bin_2', 'Bin_3', 'Bin_4']
     data_binned_equal_width = pd.cut(data_with_duplicates, bins=bins, labels=labels)

     print(f"等宽分箱后数据样本数量：{len(data_binned_equal_width)}")

可视化分箱结果

     import matplotlib.pyplot as plt

     plt.hist(data_binned_unique, bins=len(labels), edgecolor='black')
     plt.title('Distribution after Binning (Unique Values)')
     plt.xlabel('Bins')
     plt.ylabel('Frequency')
     plt.show()

     plt.hist(data_binned_equal_width, bins=len(labels), edgecolor='black')
     plt.title('Distribution after Equal Width Binning')
     plt.xlabel('Bins')
     plt.ylabel('Frequency')
     plt.show()