比较DataFrames / csv并仅返回具有差异的列，包括Key值

Question

我有两个CSV文件，我正在比较并仅并排返回具有不同值的列。df1Country 1980    1981    1982    1983    1984Bermuda 0.00793 0.00687 0.00727 0.00971 0.00752Canada  9.6947  9.58952 9.20637 9.18989 9.78546Greenland   0.00791 0.00746 0.00722 0.00505 0.00799Mexico  3.72819 4.11969 4.33477 4.06414 4.18464df2Country 1980    1981    1982    1983    1984Bermuda 0.77777 0.00687 0.00727 0.00971 0.00752Canada  9.6947  9.58952 9.20637 9.18989 9.78546Greenland   0.00791 0.00746 0.00722 0.00505 0.00799Mexico  3.72819 4.11969 4.33477 4.06414 4.18464import pandas as pdimport numpy as npdf1=pd.read_csv('csv1.csv')df2=pd.read_csv('csv2.csv')def diff_pd(df1, df2):    """Identify differences between two pandas DataFrames"""    assert (df1.columns == df2.columns).all(), \        "DataFrame column names are different"    if any(df1.dtypes != df2.dtypes):        "Data Types are different, trying to convert"        df2 = df2.astype(df1.dtypes)    if df1.equals(df2):        print("Dataframes are the same")        return None    else:        # need to account for np.nan != np.nan returning True        diff_mask = (df1 != df2) & ~(df1.isnull() & df2.isnull())        ne_stacked = diff_mask.stack()        changed = ne_stacked[ne_stacked]        changed.index.names = ['Country', 'Column']        difference_locations = np.where(diff_mask)        changed_from = df1.values[difference_locations][0]        changed_to = df2.values[difference_locations]        y=pd.DataFrame({'From': changed_from, 'To': changed_to},                            index=changed.index)        print(y)        return pd.DataFrame({'From': changed_from, 'To': changed_to},                            index=changed.index)diff_pd(df1,df2)我当前的输出是：                   From       ToCountry Column                  0       1980    0.00793  0.77777因此，我想获得索引值不匹配的行的国家/地区名称，而不是索引0。下面是一个例子。我希望我的输出是：                   From       ToCountry Column                  Bermuda  1980    0.00793  0.77777谢谢所有能提供解决方案的人。

Answer 1

一种更短的方法，在此过程中会重命名：

def process_df(df):

    res = df.set_index('Country').stack()

    res.index.rename('Column', level=1, inplace=True)

    return res

df1 = process_df(df1)

df2 = process_df(df2)

mask = (df1 != df2) & ~(df1.isnull() & df2.isnull())

df3 = pd.concat([df1[mask], df2[mask]], axis=1).rename({0:'From', 1:'To'}, axis=1)

df3

                   From       To

Country Column                  

Bermuda 1980    0.00793  0.77777