在 Pandas 中为数据框添加自定义标题并将其转换为 HTML答案

【问题标题】：Add custom title to a data-frame in Pandas and convert it to HTML在 Pandas 中为数据框添加自定义标题并将其转换为 HTML
【发布时间】：2019-10-12 04:11:01
【问题描述】：

我正在从目录列表中读取某些 csv 文件，即实际结果和预期结果。现在我遍历actual_results 中的每个csv 并比较它在expected_results 中的csvs。然后我想将整个数据显示为 HTML，如下所示

我已经编写了一些代码来实际清理数据，然后比较实际和预期 csv 的数据帧。

这是整个代码：

import pandas as pd
import sys
from glob import glob
import os
import itertools

# compareCSV takes in two args as path of the two csv files to compare
def compare(expectedList,actualList):
    ctr=0
    dfList = list()
    for (csv1,csv2) in itertools.zip_longest(expectedList,actualList):
        df1_ctr=pd.read_csv(csv1,sep=',')
        df1_ctr[df1_ctr.columns[1:]] = [x.split('\t') for x in df1_ctr['mean(ms)']]
        df1=df1_ctr.apply(pd.to_numeric,errors='coerce')
        df2_ctr=pd.read_csv(csv2,sep=',')
        df2_ctr[df2_ctr.columns[1:]] = [x.split('\t') for x in df2_ctr['mean(ms)']]
        df2=df2_ctr.apply(pd.to_numeric,errors='coerce')
        print("Dataframe for Expected List for file : {} is \n {}".format(csv1,df1))
        print("Dataframe for Actual List for file: {} is \n {}".format(csv2,df2))
        d3=df1.loc[:,:] # Dataframe 1
        d4=df2.loc[:,:] # Dataframe 2
        d5=abs(((d3.subtract(d4))/d3)*100)
        print("Deviation between file {} and {} is :\n {}".format(csv1,csv2,d5))
        ctr=ctr+1  
        #Final Data frame
        df=pd.concat([df1,df2,d5])
        #print("{}".format(df))
        dfList.append(df)   
    #print("Final Data frame: \n{}".format(dfList))
    # for data in dfList:
    #     print("data at index: \n{}".format(data))


if __name__ == "__main__":
    #file1=sys.argv[1]  # FileName1
    #file2=sys.argv[2]  #FileName2
    #compareCSV(file1,file2) # Compare CSV files passed in as paramters
    os.chdir("expected_results")
    expectedCSVs=glob("*.csv")
    #print(expectedCSVs)
    os.chdir("../actual_results")
    actualCSVs=glob("*.csv")
    #print(actualCSVs)
    compare(expectedCSVs,actualCSVs)

目前我有一些多余的打印语句。上述代码的输出如下：

Dataframe for Expected List for file : CT_QRW_25.csv is 
    100%Q   mean(ms)   P50(ms)     P99(ms)   p99.9(ms)  #Samples
0    NaN   0.038973  0.044939    0.091076    0.363859   1760108
1    NaN   0.050652  0.044963    0.094738    0.402525   1354233
2    NaN   0.046500  0.045020    0.108138    0.320636    123448
3    NaN   1.872630  0.599966   33.313200  172.040000  21954617
4    NaN  37.752900  0.600484  603.063000  805.340000   2708258
Dataframe for Actual List for file: CT_QRW_25.csv is 
    100%Q   mean(ms)   P50(ms)     P99(ms)   p99.9(ms)  #Samples
0    NaN   0.038973  0.044939    0.091076    0.363859   1760108
1    NaN   0.050652  0.044963    0.094738    0.402525   1354233
2    NaN   0.046500  0.045020    0.108138    0.320636    123448
3    NaN   1.872630  0.599966   33.313200  172.040000  21954617
4    NaN  37.752900  0.600484  603.063000  805.340000   2708258
Deviation between file CT_QRW_25.csv and CT_QRW_25.csv is :
    100%Q  mean(ms)  P50(ms)  P99(ms)  p99.9(ms)  #Samples
0    NaN       0.0      0.0      0.0        0.0       0.0
1    NaN       0.0      0.0      0.0        0.0       0.0
2    NaN       0.0      0.0      0.0        0.0       0.0
3    NaN       0.0      0.0      0.0        0.0       0.0
4    NaN       0.0      0.0      0.0        0.0       0.0
Dataframe for Expected List for file : CT_W_14.csv is 
    100%Q  mean(ms)  P50(ms)  P99(ms)  p99.9(ms)   #Samples
0    NaN       NaN      NaN      NaN        NaN        NaN
1    NaN       NaN      NaN      NaN        NaN        NaN
2    NaN       NaN      NaN      NaN        NaN        NaN
3    NaN       NaN      NaN      NaN        NaN        NaN
4    NaN   97.8025  17.8492  725.619    891.455  5304765.0
Dataframe for Actual List for file: CT_W_14.csv is 
    100%Q  mean(ms)  P50(ms)  P99(ms)  p99.9(ms)   #Samples
0    NaN       NaN      NaN      NaN        NaN        NaN
1    NaN       NaN      NaN      NaN        NaN        NaN
2    NaN       NaN      NaN      NaN        NaN        NaN
3    NaN       NaN      NaN      NaN        NaN        NaN
4    NaN   97.8025  17.8492  725.619    891.455  5304765.0
Deviation between file CT_W_14.csv and CT_W_14.csv is :
    100%Q  mean(ms)  P50(ms)  P99(ms)  p99.9(ms)  #Samples
0    NaN       NaN      NaN      NaN        NaN       NaN
1    NaN       NaN      NaN      NaN        NaN       NaN
2    NaN       NaN      NaN      NaN        NaN       NaN
3    NaN       NaN      NaN      NaN        NaN       NaN
4    NaN       0.0      0.0      0.0        0.0       0.0

目标：由于我目前拥有的语句是打印语句，因此如果我想将其转换为 HTML，我将无法使其动态化。我的目标是将其作为 HTML 文件输出。或者即使有一种自定义方法可以在数据框中添加一行作为标题，那么也可以。如果偏差大于 10%，那么我想以红色显示单元格。如果有人遇到过这种情况，那就太好了，请帮助我。任何帮助将不胜感激。

【问题讨论】：

标签： html python-3.x pandas csv data-science

【解决方案1】：

Pandas 有一个特殊的 styling 对象，可以使用 .render 方法导出到 HTML 或使用 .to_excel 导出到 Excel。您可以使用 CSS 来格式化表格并添加标题：

def highlight_high(series, threshold, colour):
    return ['background-color:'+ colour.lower() if threshold <= i else 'background-color: white' for i in series]

# df.style.apply creates a pandas.io.formats.style.Styler object from a DataFrame
highlighted = df.style.apply(highlight_high, axis=0, subset=pd.IndexSlice[:,'P50(ms)'], colour = 'red', threshold = 0.5)
# adding a caption
highlighted = highlighted.set_caption('Highlighted P50')
# render() generates the HTML for the Styler object
with open('table.html', 'w') as f:
    f.write(highlighted.render())

我不知道该涂什么颜色，所以我选择了你的Dataframe for Actual List。 Styler.apply 用于系列/数据帧样式，Styler.applymap 用于元素样式。样式函数的输入和输出的形状必须匹配。

使用pd.IndexSlice 选择带有subset 的切片
用threshold设置阈值
用colour选择HTML颜色
添加带有.set_caption 的标题
导出为 HTML .render 或 .to_excel

我的结果：

【讨论】：