Attribution Error in plotting media insights

[mahewashabdi]

Hi Team,

I successfully ran the model. however while implementing the media insights, I am facing error.

AttributeError Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_29796\1575169758.py in
----> 1 plot.plot_response_curves(media_mix_model=mmm, media_scaler=media_scaler, target_scaler=target_scaler)

~\Anaconda3\envs\lightweightmmm\lib\site-packages\lightweight_mmm\plot.py in plot_response_curves(media_mix_model, media_scaler, target_scaler, prices, optimal_allocation_per_timeunit, steps, percentage_add, apply_log_scale, figure_size, n_columns, marker_size, legend_fontsize, seed)
225 media_ranges = jnp.squeeze(media_ranges)
226 if target_scaler:
--> 227 predictions = target_scaler.inverse_transform(predictions)
228
229 if media_scaler:

~\Anaconda3\envs\lightweightmmm\lib\site-packages\lightweight_mmm\preprocessing.py in inverse_transform(self, data)
154 Dataset with the inverse transformation applied.
155 """
--> 156 return self.divide_by * data / self.multiply_by

AttributeError: 'CustomScaler' object has no attribute 'divide_by

thanks,
Mahewash

[pabloduque0]

Hello @mahewashabdi !

It looks like your target scaler has not been fit yet.

If you are not using a target scaling you dont need to pass it there, but if you are using it, make sure you are using it correctly and it has been fit before :)

Let us know if that is the case or if its something else :)

[mahewashabdi]

Hi @pabloduque0

Thanks for your reply.

I was using the target scaler. I don't know what went wrong, Its running now.. Thanks :)