Update

synthesized-io · Dec 15, 2023 · 2f96606 · 2f96606
1 parent 165ce39
commit 2f96606
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 1 deletion.
diff --git a/postgres/soda/checks_for_masking.yaml b/postgres/soda/checks_for_masking.yaml
@@ -7,7 +7,7 @@ checks for film:
   - row_count same as film in input_db
   - distribution_difference(rating) < 0.3:
       method: chi_square
-      distribution reference file: /sodacl/film_rating_distribution.yml
+      distribution reference file: /sodacl/masked_film_rating_distribution.yml
 
 checks for actor:
   - row_count same as actor in input_db

diff --git a/postgres/soda/masked_film_rating_distribution.yml b/postgres/soda/masked_film_rating_distribution.yml
@@ -0,0 +1,16 @@
+dataset: film
+column: rating
+distribution_type: categorical
+distribution_reference:
+  weights:
+  - 0.223
+  - 0.21
+  - 0.195
+  - 0.194
+  - 0.178
+  bins:
+  - QZ-54
+  - FX-15
+  - Z
+  - QZ
+  - F