Dtype object is not a valid JAX array type. Only arrays of numeric types are supported by JAX.

[EldadWinter]

When trying to do this:
media_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean)
target_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean)
cost_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean)

media_data_train = media_scaler.fit_transform(media_data_train)
target_train = target_scaler.fit_transform(target_train)
costs2 = cost_scaler.fit_transform(costs)

I got the error. up until then everything went as in the tutorial.

[entzyeung]

have you transformed your dataframe or dataset to array, e.g. with np.array(media_data_train)?

[EldadWinter]

I did like this, on the tutorial. it is an array but still doesn't let me use the Scalers:

[entzyeung]

it is an array, and is float32, it should be in good shape already. WHich dataset threw an error? target? media? or others?

[EldadWinter]

As you can see below the target_train caused errors. The Scalers & fit_transofrm of media_data_train and costs worked.

And this is how target_train looks like:

[entzyeung]

Hmmmm.... I know you have checked it already, but would there be any chance that your target_train has nan, null, zero, or different data dtype?

Do you mind to provide your notebook, and data for further investigation? I am not sure if there is any pm function in github.