Fixed some things in domain.

biotz · Jun 22, 2019 · f0805ef · f0805ef
1 parent 2303c31
commit f0805ef
Show file tree

Hide file tree

Showing 3 changed files with 5 additions and 19 deletions.
diff --git a/domain/domain.py b/domain/domain.py
@@ -226,8 +226,10 @@ def generate_domain(self):
                     # point to run inference on it since we cannot even generate
                     # a random domain. Therefore, we just ignore it from the
                     # final tensor.
-                    # if len(rand_dom_values) == 0:
-                    #     continue
+                    # We do not drop NULL cells since we stil have to repair them
+                    # with their 1 domain value.
+                    if init_value != NULL_REPR and len(rand_dom_values) == 0:
+                        continue
 
                     # Otherwise, just add the random domain values to the domain
                     # and set the cell status accordingly.

diff --git a/domain/estimators/tuple_embedding.py b/domain/estimators/tuple_embedding.py
@@ -642,18 +642,6 @@ def __init__(self, env, dataset, domain_df,
                     max(list(map(len, numerical_attr_groups)) or [0]),
                     self._embed_size)
             raise Exception()
-        # Convert non numerical init values in numerical attributes with _nan_.
-        # if self._numerical_attrs is not None:
-        #     fil_attr = self.domain_df['attribute'].isin(self._numerical_attrs)
-        #     fil_notnull = self.domain_df['weak_label'] != NULL_REPR
-        #     fil_notnumeric = self.domain_df['weak_label'].str.contains(NONNUMERICS)
-        #     bad_numerics = fil_attr & fil_notnull & fil_notnumeric
-        #     if bad_numerics.sum():
-        #         self.domain_df.loc[bad_numerics, 'weak_label'] = NULL_REPR
-        #         logging.warning('%s: replaced %d non-numerical values in DOMAIN as "%s" (NULL)',
-        #                 type(self).__name__,
-        #                 bad_numerics.sum(),
-        #                 NULL_REPR)
         # Remove domain for numerical attributes.
         fil_numattr = self.domain_df['attribute'].isin(self._numerical_attrs)
 
@@ -739,9 +727,6 @@ def __init__(self, env, dataset, domain_df,
         self.in_num_w1 = torch.nn.Parameter(torch.zeros(self._n_num_attr_groups, self._embed_size, self._embed_size))
         self.in_num_bias1 = torch.nn.Parameter(torch.zeros(self._n_num_attr_groups, self._embed_size))
 
-        # out_num_zeros_vecs may not be necessary
-        self.out_num_zero_vecs = torch.nn.Parameter(torch.zeros(self._n_train_num_attrs, self._embed_size))
-
         self.out_num_bases = torch.nn.Parameter(torch.zeros(self._n_train_num_attrs, self._embed_size, self._max_num_dim))
         # Non-linearity for combined_init for each numerical attr
         self.out_num_w1 = torch.nn.Parameter(torch.zeros(self._n_train_num_attrs, self._embed_size, self._embed_size))
@@ -783,7 +768,6 @@ def __init__(self, env, dataset, domain_df,
             torch.nn.init.xavier_uniform_(self.in_num_bias1)
 
         if self._n_train_num_attrs > 0:
-            torch.nn.init.xavier_uniform_(self.out_num_zero_vecs)
             torch.nn.init.xavier_uniform_(self.out_num_bases)
             torch.nn.init.xavier_uniform_(self.out_num_w1)
             torch.nn.init.xavier_uniform_(self.out_num_bias1)

diff --git a/tests/start_test.sh b/tests/start_test.sh
@@ -5,4 +5,4 @@ source ../set_env.sh
 
 # Launch tests.
 echo "Launching tests..."
-pytest -n 1
+pytest