MethodError: Cannot `convert` an object of type String7 to an object of type Float32

Hi,
I am writing a custom environment for a Reinforcement learning project. At every Timestep some Data is used to determine a part of the state/next state. I want to put this data in with a CSV file, but when I want to determine the next state via the step function I always get this Error:
MethodError: Cannot convert an object of type String7 to an object of type Float32

Problem in df

the weird thing is, that elektkwh went through, but only the second entry is causing an error, even if the value should be a zero…do you have any idea what I did wrong?
Thanks in advance!

Hi,

if you check the original source CSV you may find that the “heatingkwh” column is an empty string, and the env.state.d_fh is of type Float32 in the env.state struct that you are writing too.

Regards,