From dbbe0e3a6ebd2f22c386c13a3cc53184e531f837 Mon Sep 17 00:00:00 2001
From: Alex Senger <91055000+alex-senger@users.noreply.github.com>
Date: Fri, 23 Jun 2023 16:23:03 +0200
Subject: [PATCH] feat: improve `table.summary`. Catch `ValueError` thrown by
 `column.stability` (#390)

Closes #320

### Summary of Changes

The stability of a column with only null values should just be skipped
and displayed in the returned table as a minus: -

<!-- Please provide a summary of changes in this pull request, ensuring
all changes are explained. -->

---------

Co-authored-by: patrikguempel <128832338+patrikguempel@users.noreply.github.com>
Co-authored-by: megalinter-bot <129584137+megalinter-bot@users.noreply.github.com>
---
 src/safeds/data/tabular/containers/_column.py |  6 +++++
 src/safeds/data/tabular/containers/_table.py  |  2 +-
 .../containers/_column/test_stability.py      | 10 +++++++--
 .../tabular/containers/_table/test_summary.py | 22 ++++++++++++++++++-
 4 files changed, 36 insertions(+), 4 deletions(-)

diff --git a/src/safeds/data/tabular/containers/_column.py b/src/safeds/data/tabular/containers/_column.py
index ae42ace06..a4f198f06 100644
--- a/src/safeds/data/tabular/containers/_column.py
+++ b/src/safeds/data/tabular/containers/_column.py
@@ -504,6 +504,8 @@ def stability(self) -> float:
         \frac{\text{number of occurrences of most common non-null value}}{\text{number of non-null values}}
         $$
 
+        The stability cannot be calculated for a column with only null values.
+
         Returns
         -------
         stability : float
@@ -516,6 +518,10 @@ def stability(self) -> float:
         """
         if self._data.size == 0:
             raise ColumnSizeError("> 0", "0")
+
+        if self.all(lambda x: x is None):
+            raise ValueError("Stability cannot be calculated for a column with only null values.")
+
         return self._data.value_counts()[self.mode()[0]] / self._data.count()
 
     def standard_deviation(self) -> float:
diff --git a/src/safeds/data/tabular/containers/_table.py b/src/safeds/data/tabular/containers/_table.py
index b62423cab..e07117b3c 100644
--- a/src/safeds/data/tabular/containers/_table.py
+++ b/src/safeds/data/tabular/containers/_table.py
@@ -600,7 +600,7 @@ def summary(self) -> Table:
             for function in statistics.values():
                 try:
                     values.append(str(function()))
-                except NonNumericColumnError:
+                except (NonNumericColumnError, ValueError):
                     values.append("-")
 
             result = pd.concat([result, pd.DataFrame(values)], axis=1)
diff --git a/tests/safeds/data/tabular/containers/_column/test_stability.py b/tests/safeds/data/tabular/containers/_column/test_stability.py
index d1bb893d2..7de0050d9 100644
--- a/tests/safeds/data/tabular/containers/_column/test_stability.py
+++ b/tests/safeds/data/tabular/containers/_column/test_stability.py
@@ -25,7 +25,13 @@ def test_should_return_stability_of_column(values: list[Any], expected: float) -
     assert column.stability() == expected
 
 
-def test_should_raise_if_column_is_empty() -> None:
+def test_should_raise_column_size_error_if_column_is_empty() -> None:
     column: Column[Any] = Column("A", [])
-    with pytest.raises(ColumnSizeError):
+    with pytest.raises(ColumnSizeError, match="Expected a column of size > 0 but got column of size 0."):
+        column.stability()
+
+
+def test_should_raise_value_error_if_column_contains_only_none() -> None:
+    column: Column[Any] = Column("A", [None, None])
+    with pytest.raises(ValueError, match="Stability cannot be calculated for a column with only null values."):
         column.stability()
diff --git a/tests/safeds/data/tabular/containers/_table/test_summary.py b/tests/safeds/data/tabular/containers/_table/test_summary.py
index 460bc287c..f4abaa988 100644
--- a/tests/safeds/data/tabular/containers/_table/test_summary.py
+++ b/tests/safeds/data/tabular/containers/_table/test_summary.py
@@ -112,8 +112,28 @@
                 },
             ),
         ),
+        (
+            Table({"col": [None, None]}),
+            Table(
+                {
+                    "metrics": [
+                        "maximum",
+                        "minimum",
+                        "mean",
+                        "mode",
+                        "median",
+                        "sum",
+                        "variance",
+                        "standard deviation",
+                        "idness",
+                        "stability",
+                    ],
+                    "col": ["-", "-", "-", "[]", "-", "-", "-", "-", "0.0", "-"],
+                },
+            ),
+        ),
     ],
-    ids=["Column of integers and Column of characters", "empty", "empty with columns"],
+    ids=["Column of integers and Column of characters", "empty", "empty with columns", "Column of None"],
 )
 def test_should_make_summary(table: Table, expected: Table) -> None:
     assert expected.schema == table.summary().schema