I have a pyspark dataframe and try to create a sequence with customized intervals. These timesteps are stored in the column "new_timestep" as double value (but I also
