About DMLIV new release

Hi,

EDITED:
I had issues on store_final and inference.
However I tried:

cate = DMLIV(model_Y_X =model_Y_X(), model_T_X =model_T_X(), model_T_XZ =model_T_XZ(),
             model_final=dmliv_model_effect(), featurizer=dmliv_featurizer(),cv=6,
             n_splits= 20)

and seems to work with store_final = cache_values. Further, I saw from the documentation that it can be specified inference= 'bootstrap'` when fitting. Two main issues emerge:

1. specifying inference= 'bootstrap' in cate.fit() takes forever to run. Also without specifying it, it takes a lot of time and the following error emerges:
   The final model has a nonzero intercept for at least one outcome; it will be subtracted, but consider fitting a model without an intercept if possible.
   I guess this is maybe due to the fact that the panel wrapper provided in previous issues does not exclude the first time dummy and individual dummy when performing:

n_products_per_group = 1
n_months = 47
n_groups = 11
n_products = n_groups * n_products_per_group
# p[i, j] = price of product i on month j
p = price
# p_minus[i, j] = average price of products other than i within the same group on month j
p_minus = med_price_atc
# q[i, j] = demand of product i on month j
q = demand
# X[i, j] = (one-hot-encoding of product i): si fa - 3 perchè sarebbe nprod-1+nmonth-1+group-1
X = np.zeros((n_products, n_months, n_products + n_months + n_groups - 2))

for g in np.arange(n_groups):
    for i in np.arange(n_products_per_group):
        for j in np.arange(n_months):
            index_i = n_products_per_group*g + i
            X[index_i, j, :(n_products)] = 1.0 * (np.arange(0, n_products) == (index_i))
            #INDICE GRUPPO
            X[index_i, j, (n_products):(n_products + n_groups - 1)] =  1.0 * (np.arange(1, n_groups) == g)
            #INDICE QUARTER
            X[index_i, j, (n_products + n_groups - 1):] = 1.0 * (np.arange(1, n_months) == j)

Could you please confirm that?

2. is it possible to use inference='bootstrap' with cate.refit_final() ?

3. I cannot import SubsetWrapper, where can I get it from using the latest release?

Thank you very much!

Start by reproducing the shown DMLIV configuration, including cate.fit(), cate.refit_final(), bootstrap inference, and the attempted SubsetWrapper import. Check the latest release documentation and package exports for these entry points. Done means determining whether the reported runtime, intercept warning, refit behavior, and import failure are current reproducible bugs or release-specific usage issues.