From 5f002db4661f8d637312862b5aabaed19e43fb46 Mon Sep 17 00:00:00 2001
From: Terji Petersen <terjipetersen@Terjis-Air.fritz.box>
Date: Sat, 19 Nov 2022 11:38:25 +0000
Subject: [PATCH 1/3] TST: add tests/arrays/interval/test_constructors.py

---
 .../arrays/interval/test_constructors.py      | 483 ++++++++++++++++++
 1 file changed, 483 insertions(+)
 create mode 100644 pandas/tests/arrays/interval/test_constructors.py

diff --git a/pandas/tests/arrays/interval/test_constructors.py b/pandas/tests/arrays/interval/test_constructors.py
new file mode 100644
index 0000000000000..bc2cf439882a9
--- /dev/null
+++ b/pandas/tests/arrays/interval/test_constructors.py
@@ -0,0 +1,483 @@
+from functools import partial
+
+import numpy as np
+import pytest
+
+from pandas.core.dtypes.common import is_categorical_dtype
+from pandas.core.dtypes.dtypes import IntervalDtype
+
+from pandas import (
+    Categorical,
+    CategoricalIndex,
+    Index,
+    Interval,
+    Series,
+    date_range,
+    notna,
+    period_range,
+    timedelta_range,
+)
+import pandas._testing as tm
+from pandas.api.types import is_unsigned_integer_dtype
+from pandas.core.api import (
+    Float64Index,
+    Int64Index,
+    UInt64Index,
+)
+from pandas.core.arrays import IntervalArray
+import pandas.core.common as com
+
+
+@pytest.fixture(params=[None, "foo"])
+def name(request):
+    return request.param
+
+
+class ConstructorTests:
+    """
+    Common tests for all variations of IntervalArray construction. Input data
+    to be supplied in breaks format, then converted by the subclass method
+    get_kwargs_from_breaks to the expected format.
+    """
+
+    # get_kwargs_from_breaks in TestFromTuples and TestClassconstructors just return
+    # tuples of ints, so IntervalArray can't know the original dtype was uint
+    _use_dtype_in_test_constructor_uint = False
+
+    @pytest.mark.parametrize(
+        "breaks",
+        [
+            [3, 14, 15, 92, 653],
+            np.arange(10, dtype="int64"),
+            np.arange(10, dtype="uint64"),
+            np.arange(20, 30, 0.5),
+            date_range("20180101", periods=10),
+            date_range("20180101", periods=10, tz="US/Eastern"),
+            timedelta_range("1 day", periods=10),
+        ],
+    )
+    @pytest.mark.parametrize("use_dtype", [True, False])
+    def test_constructor(self, constructor, breaks, closed, use_dtype):
+        breaks_dtype = getattr(breaks, "dtype", "int64")
+        result_kwargs = self.get_kwargs_from_breaks(breaks, closed)
+        is_uint = is_unsigned_integer_dtype(breaks_dtype)
+        if use_dtype or (self._use_dtype_in_test_constructor_uint and is_uint):
+            result_kwargs["dtype"] = IntervalDtype(breaks_dtype, closed=closed)
+
+        result = constructor(closed=closed, **result_kwargs)
+
+        assert result.closed == closed
+        assert result.dtype.subtype == breaks_dtype
+        tm.assert_index_equal(result.left, Index(breaks[:-1]))
+        tm.assert_index_equal(result.right, Index(breaks[1:]))
+
+    @pytest.mark.parametrize(
+        "breaks, subtype",
+        [
+            (Int64Index([0, 1, 2, 3, 4]), "float64"),
+            (Int64Index([0, 1, 2, 3, 4]), "datetime64[ns]"),
+            (Int64Index([0, 1, 2, 3, 4]), "timedelta64[ns]"),
+            (Float64Index([0, 1, 2, 3, 4]), "int64"),
+            (date_range("2017-01-01", periods=5), "int64"),
+            (timedelta_range("1 day", periods=5), "int64"),
+        ],
+    )
+    def test_constructor_dtype(self, constructor, breaks, subtype):
+        # GH 19262: conversion via dtype parameter
+        expected_kwargs = self.get_kwargs_from_breaks(breaks.astype(subtype))
+        expected = constructor(**expected_kwargs)
+
+        result_kwargs = self.get_kwargs_from_breaks(breaks)
+        iv_dtype = IntervalDtype(subtype, "right")
+        for dtype in (iv_dtype, str(iv_dtype)):
+            result = constructor(dtype=dtype, **result_kwargs)
+            tm.assert_interval_array_equal(result, expected)
+
+    @pytest.mark.parametrize(
+        "breaks",
+        [
+            Int64Index([0, 1, 2, 3, 4]),
+            UInt64Index([0, 1, 2, 3, 4]),
+            Float64Index([0, 1, 2, 3, 4]),
+            date_range("2017-01-01", periods=5),
+            timedelta_range("1 day", periods=5),
+        ],
+    )
+    def test_constructor_pass_closed(self, constructor, breaks):
+        # not passing closed to IntervalDtype, but to IntervalArray constructor
+        iv_dtype = IntervalDtype(breaks.dtype)
+
+        result_kwargs = self.get_kwargs_from_breaks(breaks)
+
+        for dtype in (iv_dtype, str(iv_dtype)):
+            with tm.assert_produces_warning(None):
+
+                result = constructor(dtype=dtype, closed="left", **result_kwargs)
+            assert result.dtype.closed == "left"
+
+    @pytest.mark.parametrize("breaks", [[np.nan] * 2, [np.nan] * 4, [np.nan] * 50])
+    def test_constructor_nan(self, constructor, breaks, closed):
+        # GH 18421
+        result_kwargs = self.get_kwargs_from_breaks(breaks)
+        result = constructor(closed=closed, **result_kwargs)
+
+        expected_subtype = np.float64
+        expected_values = np.array(breaks[:-1], dtype=object)
+
+        assert result.closed == closed
+        assert result.dtype.subtype == expected_subtype
+        tm.assert_numpy_array_equal(np.array(result), expected_values)
+
+    @pytest.mark.parametrize(
+        "breaks",
+        [
+            [],
+            np.array([], dtype="int64"),
+            np.array([], dtype="uint64"),
+            np.array([], dtype="float64"),
+            np.array([], dtype="datetime64[ns]"),
+            np.array([], dtype="timedelta64[ns]"),
+        ],
+    )
+    def test_constructor_empty(self, constructor, breaks, closed):
+        # GH 18421
+        result_kwargs = self.get_kwargs_from_breaks(breaks)
+        result = constructor(closed=closed, **result_kwargs)
+
+        expected_values = np.array([], dtype=object)
+        expected_subtype = getattr(breaks, "dtype", np.int64)
+
+        assert len(result) == 0
+        assert result.closed == closed
+        assert result.dtype.subtype == expected_subtype
+        tm.assert_numpy_array_equal(np.array(result), expected_values)
+
+    @pytest.mark.parametrize(
+        "breaks",
+        [
+            tuple("0123456789"),
+            list("abcdefghij"),
+            np.array(list("abcdefghij"), dtype=object),
+            np.array(list("abcdefghij"), dtype="<U1"),
+        ],
+    )
+    def test_constructor_string(self, constructor, breaks):
+        # GH 19016
+        msg = (
+            "category, object, and string subtypes are not supported "
+            "for IntervalArray"
+        )
+        with pytest.raises(TypeError, match=msg):
+            constructor(**self.get_kwargs_from_breaks(breaks))
+
+    @pytest.mark.parametrize("cat_constructor", [Categorical, CategoricalIndex])
+    def test_constructor_categorical_valid(self, constructor, cat_constructor):
+        # GH 21243/21253
+
+        breaks = np.arange(10, dtype="int64")
+        expected = IntervalArray.from_breaks(breaks)
+
+        cat_breaks = cat_constructor(breaks)
+        result_kwargs = self.get_kwargs_from_breaks(cat_breaks)
+        result = constructor(**result_kwargs)
+        tm.assert_interval_array_equal(result, expected)
+
+    def test_generic_errors(self, constructor):
+        # filler input data to be used when supplying invalid kwargs
+        filler = self.get_kwargs_from_breaks(range(10))
+
+        # invalid closed
+        msg = "closed must be one of 'right', 'left', 'both', 'neither'"
+        with pytest.raises(ValueError, match=msg):
+            constructor(closed="invalid", **filler)
+
+        # unsupported dtype
+        msg = "dtype must be an IntervalDtype, got int64"
+        with pytest.raises(TypeError, match=msg):
+            constructor(dtype="int64", **filler)
+
+        # invalid dtype
+        msg = "data type [\"']invalid[\"'] not understood"
+        with pytest.raises(TypeError, match=msg):
+            constructor(dtype="invalid", **filler)
+
+        # no point in nesting periods in an IntervalArray
+        periods = period_range("2000-01-01", periods=10)
+        periods_kwargs = self.get_kwargs_from_breaks(periods)
+        msg = "Period dtypes are not supported, use a PeriodIndex instead"
+        with pytest.raises(ValueError, match=msg):
+            constructor(**periods_kwargs)
+
+        # decreasing values
+        decreasing_kwargs = self.get_kwargs_from_breaks(range(10, -1, -1))
+        msg = "left side of interval must be <= right side"
+        with pytest.raises(ValueError, match=msg):
+            constructor(**decreasing_kwargs)
+
+
+class TestFromArrays(ConstructorTests):
+    """Tests specific to IntervalArray.from_arrays"""
+
+    @pytest.fixture
+    def constructor(self):
+        return IntervalArray.from_arrays
+
+    def get_kwargs_from_breaks(self, breaks, closed="right"):
+        """
+        converts intervals in breaks format to a dictionary of kwargs to
+        specific to the format expected by IntervalArray.from_arrays
+        """
+        return {"left": breaks[:-1], "right": breaks[1:]}
+
+    def test_constructor_errors(self):
+        # GH 19016: categorical data
+        data = Categorical(list("01234abcde"), ordered=True)
+        msg = (
+            "category, object, and string subtypes are not supported "
+            "for IntervalArray"
+        )
+        with pytest.raises(TypeError, match=msg):
+            IntervalArray.from_arrays(data[:-1], data[1:])
+
+        # unequal length
+        left = [0, 1, 2]
+        right = [2, 3]
+        msg = "left and right must have the same length"
+        with pytest.raises(ValueError, match=msg):
+            IntervalArray.from_arrays(left, right)
+
+    @pytest.mark.parametrize(
+        "left_subtype, right_subtype", [(np.int64, np.float64), (np.float64, np.int64)]
+    )
+    def test_mixed_float_int(self, left_subtype, right_subtype):
+        """mixed int/float left/right results in float for both sides"""
+        left = np.arange(9, dtype=left_subtype)
+        right = np.arange(1, 10, dtype=right_subtype)
+        result = IntervalArray.from_arrays(left, right)
+
+        expected_left = Float64Index(left)
+        expected_right = Float64Index(right)
+        expected_subtype = np.float64
+
+        tm.assert_index_equal(result.left, expected_left)
+        tm.assert_index_equal(result.right, expected_right)
+        assert result.dtype.subtype == expected_subtype
+
+
+class TestFromBreaks(ConstructorTests):
+    """Tests specific to IntervalArray.from_breaks"""
+
+    @pytest.fixture
+    def constructor(self):
+        return IntervalArray.from_breaks
+
+    def get_kwargs_from_breaks(self, breaks, closed="right"):
+        """
+        converts intervals in breaks format to a dictionary of kwargs to
+        specific to the format expected by IntervalArray.from_breaks
+        """
+        return {"breaks": breaks}
+
+    def test_constructor_errors(self):
+        # GH 19016: categorical data
+        data = Categorical(list("01234abcde"), ordered=True)
+        msg = (
+            "category, object, and string subtypes are not supported "
+            "for IntervalArray"
+        )
+        with pytest.raises(TypeError, match=msg):
+            IntervalArray.from_breaks(data)
+
+    def test_length_one(self):
+        """breaks of length one produce an empty IntervalArray"""
+        """breaks of length one produce an empty IntervalArray"""
+        breaks = [0]
+        result = IntervalArray.from_breaks(breaks)
+        expected = IntervalArray.from_breaks([])
+        tm.assert_interval_array_equal(result, expected)
+
+    def test_left_right_dont_share_data(self):
+        # GH#36310
+        breaks = np.arange(5)
+        result = IntervalArray.from_breaks(breaks)
+        assert result._left.base is None or result._left.base is not result._right.base
+
+
+class TestFromTuples(ConstructorTests):
+    """Tests specific to IntervalArray.from_tuples"""
+
+    _use_dtype_in_test_constructor_uint = True
+
+    @pytest.fixture
+    def constructor(self):
+        return IntervalArray.from_tuples
+
+    def get_kwargs_from_breaks(self, breaks, closed="right"):
+        """
+        converts intervals in breaks format to a dictionary of kwargs to
+        specific to the format expected by IntervalArray.from_tuples
+        """
+        if len(breaks) == 0:
+            return {"data": breaks}
+
+        tuples = list(zip(breaks[:-1], breaks[1:]))
+        if isinstance(breaks, (list, tuple)):
+            return {"data": tuples}
+        elif is_categorical_dtype(breaks):
+            return {"data": breaks._constructor(tuples)}
+        return {"data": com.asarray_tuplesafe(tuples)}
+
+    def test_constructor_errors(self):
+        # non-tuple
+        tuples = [(0, 1), 2, (3, 4)]
+        msg = "IntervalArray.from_tuples received an invalid item, 2"
+        with pytest.raises(TypeError, match=msg.format(t=tuples)):
+            IntervalArray.from_tuples(tuples)
+
+        # too few/many items
+        tuples = [(0, 1), (2,), (3, 4)]
+        msg = "IntervalArray.from_tuples requires tuples of length 2, got {t}"
+        with pytest.raises(ValueError, match=msg.format(t=tuples)):
+            IntervalArray.from_tuples(tuples)
+
+        tuples = [(0, 1), (2, 3, 4), (5, 6)]
+        with pytest.raises(ValueError, match=msg.format(t=tuples)):
+            IntervalArray.from_tuples(tuples)
+
+    def test_na_tuples(self):
+        # tuple (NA, NA) evaluates the same as NA as an element
+        na_tuple = [(0, 1), (np.nan, np.nan), (2, 3)]
+        idx_na_tuple = IntervalArray.from_tuples(na_tuple)
+        idx_na_element = IntervalArray.from_tuples([(0, 1), np.nan, (2, 3)])
+        tm.assert_interval_array_equal(idx_na_tuple, idx_na_element)
+
+
+class TestClassConstructors(ConstructorTests):
+    """Tests specific to the IntervalArray/Serie constructors"""
+
+    _use_dtype_in_test_constructor_uint = True
+
+    @pytest.fixture(
+        params=[IntervalArray, partial(Series, dtype="interval")],
+        ids=["IntervalArray", "Series"],
+    )
+    def klass(self, request):
+        # We use a separate fixture here to include Series with dtype kwarg
+        return request.param
+
+    @pytest.fixture
+    def constructor(self):
+        return IntervalArray
+
+    def get_kwargs_from_breaks(self, breaks, closed="right"):
+        """
+        converts intervals in breaks format to a dictionary of kwargs to
+        specific to the format expected by the IntervalArray/Series constructors
+        """
+        if len(breaks) == 0:
+            return {"data": breaks}
+
+        ivs = [
+            Interval(left, right, closed) if notna(left) else left
+            for left, right in zip(breaks[:-1], breaks[1:])
+        ]
+
+        if isinstance(breaks, list):
+            return {"data": ivs}
+        elif is_categorical_dtype(breaks):
+            return {"data": breaks._constructor(ivs)}
+        return {"data": np.array(ivs, dtype=object)}
+
+    def test_generic_errors(self, constructor):
+        """
+        override the base class implementation since errors are handled
+        differently; checks unnecessary since caught at the Interval level
+        """
+
+    def test_constructor_string(self):
+        # GH23013
+        # When forming the interval from breaks,
+        # the interval of strings is already forbidden.
+        pass
+
+    def test_constructor_errors(self, klass):
+        # mismatched closed within intervals with no constructor override
+        ivs = [Interval(0, 1, closed="right"), Interval(2, 3, closed="left")]
+        msg = "intervals must all be closed on the same side"
+        with pytest.raises(ValueError, match=msg):
+            klass(ivs)
+
+        # scalar
+        if klass is IntervalArray:
+            msg = (
+                r"IntervalArray\(...\) must be called with a collection of "
+                "some kind, 5 was passed"
+            )
+        else:
+            msg = r"type <class 'numpy.int64'> with value 5 is not an interval"
+        with pytest.raises(TypeError, match=msg):
+            klass(5)
+
+        # not an interval; dtype depends on 32bit/windows builds
+        msg = "type <class 'numpy.int(32|64)'> with value 0 is not an interval"
+        with pytest.raises(TypeError, match=msg):
+            klass([0, 1])
+
+    @pytest.mark.parametrize(
+        "data, closed",
+        [
+            ([], "both"),
+            ([np.nan, np.nan], "neither"),
+            (
+                [Interval(0, 3, closed="neither"), Interval(2, 5, closed="neither")],
+                "left",
+            ),
+            (
+                [Interval(0, 3, closed="left"), Interval(2, 5, closed="right")],
+                "neither",
+            ),
+            (IntervalArray.from_breaks(range(5), closed="both"), "right"),
+        ],
+    )
+    def test_override_inferred_closed(self, constructor, data, closed):
+        # GH 19370
+        if isinstance(data, IntervalArray):
+            tuples = data.to_tuples()
+        else:
+            tuples = [(iv.left, iv.right) if notna(iv) else iv for iv in data]
+        expected = IntervalArray.from_tuples(tuples, closed=closed)
+        result = constructor(data, closed=closed)
+        tm.assert_interval_array_equal(result, expected)
+
+    @pytest.mark.parametrize(
+        "values_constructor", [list, np.array, IntervalArray, Series]
+    )
+    def test_index_object_dtype(self, values_constructor):
+        intervals = [Interval(0, 1), Interval(1, 2), Interval(2, 3)]
+        values = values_constructor(intervals)
+        result = Series(values, dtype="interval")
+
+        assert result.dtype == "interval"
+        tm.assert_numpy_array_equal(result.values, np.array(values))
+
+    def test_array_mixed_closed(self):
+        # GH27172
+        intervals = [
+            Interval(0, 1, closed="left"),
+            Interval(1, 2, closed="right"),
+            Interval(2, 3, closed="neither"),
+            Interval(3, 4, closed="both"),
+        ]
+        msg = "intervals must all be closed on the same side"
+        with pytest.raises(ValueError, match=msg):
+            IntervalArray(intervals)
+
+
+def test_dtype_closed_mismatch():
+    # GH#38394 closed specified in both dtype and IntervalArray constructor
+
+    dtype = IntervalDtype(np.int64, "left")
+
+    msg = "closed keyword does not match dtype.closed"
+    with pytest.raises(ValueError, match=msg):
+        IntervalArray([], dtype=dtype, closed="neither")

From 87229fc47836c741b0083442d024348a3f7ff92e Mon Sep 17 00:00:00 2001
From: Terji Petersen <terjipetersen@Terjis-Air.fritz.box>
Date: Sat, 19 Nov 2022 12:17:13 +0000
Subject: [PATCH 2/3] fix spelling

---
 pandas/tests/arrays/interval/test_constructors.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/tests/arrays/interval/test_constructors.py b/pandas/tests/arrays/interval/test_constructors.py
index bc2cf439882a9..10d712b12c240 100644
--- a/pandas/tests/arrays/interval/test_constructors.py
+++ b/pandas/tests/arrays/interval/test_constructors.py
@@ -353,7 +353,7 @@ def test_na_tuples(self):
 
 
 class TestClassConstructors(ConstructorTests):
-    """Tests specific to the IntervalArray/Serie constructors"""
+    """Tests specific to the IntervalArray/Series constructors"""
 
     _use_dtype_in_test_constructor_uint = True
 

From e3d99e2c483c66a7d4cf82fcb27194da71b6cb38 Mon Sep 17 00:00:00 2001
From: Terji Petersen <terjipetersen@Terjis-Air.fritz.box>
Date: Sat, 19 Nov 2022 13:27:41 +0000
Subject: [PATCH 3/3] fix 32bit error msg

---
 pandas/tests/arrays/interval/test_constructors.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/tests/arrays/interval/test_constructors.py b/pandas/tests/arrays/interval/test_constructors.py
index 10d712b12c240..136ded26c9955 100644
--- a/pandas/tests/arrays/interval/test_constructors.py
+++ b/pandas/tests/arrays/interval/test_constructors.py
@@ -414,7 +414,7 @@ def test_constructor_errors(self, klass):
                 "some kind, 5 was passed"
             )
         else:
-            msg = r"type <class 'numpy.int64'> with value 5 is not an interval"
+            msg = "type <class 'numpy.int(32|64)'> with value 5 is not an interval"
         with pytest.raises(TypeError, match=msg):
             klass(5)