Pandas subtract 2 rows from same dataframe

问题

How do I subtract one row from another in the following dataframe (df):

RECL_LCC          1          2          3
RECL_LCC  35.107655  36.015210  28.877135
RECL_PI   36.961519  43.499506  19.538975

I want to do something like:

df['Difference'] = df['RECL_LCC']-df['RECL_PI']

but that gives:

*** KeyError: 'RECL_LCC'

回答1:

You can select rows by index value using df.loc:

In [98]: df.loc['Diff'] = df.loc['RECL_LCC'] - df.loc['RECL_PI']

In [99]: df
Out[99]: 
RECL_LCC          1          2          3
RECL_LCC  35.107655  36.015210  28.877135
RECL_PI   36.961519  43.499506  19.538975
Diff      -1.853864  -7.484296   9.338160

回答2:

you can use diff() function:

df.set_index('RECT_LCC', inplace=True)
df.diff(-1)


                    1           2         3
RECT_LCC            
RECT_LCC    -1.853864   -7.484296   9.33816
RECL_PI           NaN         NaN       NaN

by default, it shifts by 1 row. In your case, since you subtracting next row instead of previous you need to set diff(-1)

来源：https://stackoverflow.com/questions/26685600/pandas-subtract-2-rows-from-same-dataframe