熊猫在groupby内插

>>> df.groupby('filename').apply(lambda group: group.interpolate(method='index'))    filename  val1  val2t 1  file1.csv     5    102  file1.csv    10    153  file1.csv    15    206  file2.csv   NaN   NaN7  file2.csv    10    208  file2.csv    12    15