Added working test cases for duckdb

Madhu1512 · Nov 26, 2022 · fc2da4e · fc2da4e
1 parent 5dcb2a9
commit fc2da4e
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 5 deletions.
diff --git a/cuallee/duckdb_validation.py b/cuallee/duckdb_validation.py
@@ -28,7 +28,7 @@ def is_unique(self, rule: Rule) -> str:
         return f"COUNT(DISTINCT({rule.column}))"
 
     def are_unique(self, rule: Rule) -> str:
-        return f"COUNT(DISTINCT{rule.column}) / {float(len(rule.column))}"
+        return "( " + " + ".join([f"approx_count_distinct({column})" for column in rule.column]) + f") / cast({float(len(rule.column))} AS FLOAT)"
 
     def is_greater_than(self, rule: Rule) -> str:
         return f"CAST({rule.column} > {rule.value} AS INTEGER)"

diff --git a/test/unit/duckdb_dataframe/test_are_unique.py b/test/unit/duckdb_dataframe/test_are_unique.py
@@ -3,22 +3,20 @@
 import pytest
 import duckdb
 
-@pytest.mark.skip(reason="Not implemented")
 def test_positive(check: Check, db: duckdb.DuckDBPyConnection):
     check.are_unique(("id", "id2"))
     df = pd.DataFrame({"id": [10, 20], "id2": [300, 500]})
     check.table_name = "df"
     assert check.validate(db).status.str.match("PASS").all()
 
 
-@pytest.mark.skip(reason="Not implemented")
 def test_negative(check: Check, db: duckdb.DuckDBPyConnection):
     check.are_unique(("id", "id2"))
     df = pd.DataFrame({"id": [10, 10], "id2": [300, 300]})
     check.table_name = "df"
     assert check.validate(db).status.str.match("FAIL").all()
 
-@pytest.mark.skip(reason="Not implemented")
+
 @pytest.mark.parametrize(
     "rule_column", [tuple(["id", "id2"]), list(["id", "id2"])], ids=("tuple", "list")
 )
@@ -29,11 +27,13 @@ def test_parameters(check: Check, db: duckdb.DuckDBPyConnection, rule_column):
     result = check.validate(db)
     assert result.status.str.match("FAIL").all()
 
-@pytest.mark.skip(reason="Not implemented")
+
 def test_coverage(check: Check, db: duckdb.DuckDBPyConnection):
     check.are_unique(("id", "id2"), 0.75)
     df = pd.DataFrame({"id": [10, None], "id2": [300, 500]})
     check.table_name = "df"
     result = check.validate(db)
     assert result.status.str.match("PASS").all()
     assert result.pass_rate.max() == 0.75
+
+