JAVA订单报表数据清洗机制_pandas进行数据预处理(数据清洗)

前言：最近在进行NLP学习时，聚类得到的效果总是差强人意，反复思考后发现自己的数据集非常不规范，因此，学习python的pandas库对数据进行预处理。

1.数据集加载

import pandas as pd

import numpy as np

def dataSet():

data_set = [["蒜蓉生耗", "蒙古烤羊腿", "白斩鸡", "纸杯蛋糕", "蒜蓉生耗"],

["有", "无", "有", "有", "有"],

["本店招牌菜，进店必点", "店主强烈推荐", "不辣", "", "本店招牌菜，进店必点"],

[40, 340, 120, 20, 40]]

df = pd.DataFrame(np.array(data_set).T, columns=["菜名", "存货", "描述", "价格"])

df.iloc[2, 1] = np.NaN

return df

def appendSet():

data_set = [["蒜蓉生耗", "蒙古烤羊腿", "白斩鸡", "纸杯蛋糕", "蒜蓉生耗"],

[423, 300, 20, 83, 423]]

return pd.DataFrame(np.array(data_set).T, columns=["菜名",