等间距抽样

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# 预设风格
from matplotlib.pylab import style #自定义图表风格
style.use('ggplot')
# 绘图参数设置
plt.rcParams['font.sans-serif'] = ['SimHei']  # 步骤一（替换sans-serif字体）
plt.rcParams['axes.unicode_minus'] = False  # 步骤二（解决坐标轴负数的负号显示问题）
import warnings
warnings.filterwarnings("ignore")
pd.set_option('mode.chained_assignment', None)  # 直接关闭警告

def equal_interval_sampling(data, mnum):
    col = 'SEL_MW'     
    # mnum = 2000   # 样本数量
    astep = (data.max()[col] - data.min()[col]) / mnum  # 计算步长
    astep = np.round(astep, 1)
    age_bins = np.round(np.arange(data.min()[col], data.max()[col] + 1, astep), 3)  # 计算工况区间
    print(f'步长：{astep}\n区间个数：{len(age_bins) - 1}')
    ind_list = []
    for val in age_bins:
        fh = data.iloc[:, -1]
        bf = np.abs(fh - val)  
        sbf = bf.sort_values().round(3)  # 按照距离排序
        sbf = sbf.reset_index()
        sbf.columns = ['编号', '分割点']
        ind_list.append(sbf.loc[0, '编号'] )   # 保存距离最近点的索引
    print('排序前样本数目：', len(ind_list))
    end_row = sorted(set(ind_list))
    # print(end_row)
    print('排序后样本数目：', len(end_row))
    end_df = data.iloc[end_row, :]
    end_df = end_df.reset_index(drop=True)
    return end_df


from Memorymatrix import IntervalSampl 
mnum = 2000    # 等间隔抽样数量
cdata = pd.read_csv(r"a.csv", encoding='gbk')

sdata =IntervalSampl.equal_interval_sampling(cdata, mnum)
print(sdata.shape)
sdata.head(3)