Search code examples
pythonpandasdataframerapids

TypeError: melt() takes 1 positional argument but 2 were given

I am trying to use melt() function but it is showing me an error for passing 2 argument, which really weird because i am passing id as an argument and in my DataFrame i have only one id column, Although this error only comes when i use data which split from dataset by train_test_split function otherwise it is working fine.

error message:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-676-bfd6b4109c81> in <module>
      1 # melt into long form
----> 2 X_train = X_train.melt('id')
      3 
      4 # group on melted id for idxmax/mean per "row"
      5 g = X_train.groupby('id').value

TypeError: melt() takes 1 positional argument but 2 were given

here X_train is a cudf DataFrame.

Solution

  • The positional parameter is self, which is X_train. melt expects its parameters to be specified by keyword. Try X_train.melt(id_vars=['id']).