python - 如何从 pandas diff 获取单元格位置？

Question

df1 = pd.read_excel(mxln)  # Loads master xlsx for comparison
df2 = pd.read_excel(sfcn)  # Loads student xlsx for comparison
difference = df2[df2 != df1]  # Scans for differences

无论哪里有差异，我都想将这些单元格位置存储在列表中。它需要采用“A1”格式(而不是 [1, 1] 之类的格式)，以便我可以通过以下方式传递它:

redFill = PatternFill(start_color='FFEE1111', end_color='FFEE1111', fill_type='solid')
lsws['A1'].fill = redFill
lsfh.save(sfcn)  

我看过类似 this 的解决方案，但我无法让它工作/不明白它。例如，以下内容不起作用:

def highlight_cells():
    df1 = pd.read_excel(mxln)  # Loads master xlsx for comparison
    df2 = pd.read_excel(sfcn)  # Loads student xlsx for comparison
    difference = df2[df2 != df1]  # Scans for differences
    return ['background-color: yellow']

df2.style.apply(highlight_cells) 

Answer 1

从两个 pandas.DataFrame 中获取差异单元格作为 excel 坐标，您可以这样做:

代码:

def diff_cell_indices(dataframe1, dataframe2):
    from openpyxl.utils import get_column_letter as column_letter

    x_ofs = dataframe1.columns.nlevels + 1
    y_ofs = dataframe1.index.nlevels + 1
    return [column_letter(x + x_ofs) + str(y + y_ofs) for
            y, x in zip(*np.where(dataframe1 != dataframe2))]

测试代码:

import pandas as pd
df1 = pd.read_excel('test.xlsx')
print(df1)

df2 = df.copy()
df2.C['R2'] = 1
print(df2)

print(diff_cell_indices(df1, df2))

结果:

    B  C
R2  2  3
R3  4  5

    B  C
R2  2  1
R3  4  5

['C2']