resolve merge conflicts

Author: simonjayhawkins

.gitignore

@@ -57,6 +57,7 @@ dist
 # wheel files
 *.whl
 **/wheelhouse/*
+pip-wheel-metadata
 # coverage
 .coverage
 coverage.xml

MANIFEST.in

@@ -3,6 +3,7 @@ include LICENSE
 include RELEASE.md
 include README.md
 include setup.py
+include pyproject.toml
 
 graft doc
 prune doc/build

Makefile

@@ -18,7 +18,7 @@ black:
 	black . --exclude '(asv_bench/env|\.egg|\.git|\.hg|\.mypy_cache|\.nox|\.tox|\.venv|_build|buck-out|build|dist|setup.py)'
 
 develop: build
-	python setup.py develop
+	python -m pip install --no-build-isolation -e .
 
 doc:
 	-rm -rf doc/build doc/source/generated

README.md

@@ -188,16 +188,17 @@ python setup.py install
 
 or for installing in [development mode](https://pip.pypa.io/en/latest/reference/pip_install.html#editable-installs):
 
+
 ```sh
-python setup.py develop
+python -m pip install --no-build-isolation -e .
 ```
 
-Alternatively, you can use `pip` if you want all the dependencies pulled
-in automatically (the `-e` option is for installing it in [development
-mode](https://pip.pypa.io/en/latest/reference/pip_install.html#editable-installs)):
+If you have `make`, you can also use `make develop` to run the same command.
+
+or alternatively
 
 ```sh
-pip install -e .
+python setup.py develop
 ```
 
 See the full instructions for [installing from source](https://pandas.pydata.org/pandas-docs/stable/install.html#installing-from-source).

ci/incremental/build.cmd

@@ -1,9 +1,9 @@
 @rem https://github.com/numba/numba/blob/master/buildscripts/incremental/build.cmd
 
-@rem Build numba extensions without silencing compile errors
-python setup.py build_ext -q --inplace
+@rem Build extensions
+python setup.py build_ext -q -i
 
-@rem Install pandas locally
-python -m pip install -e .
+@rem Install pandas
+python -m pip install --no-build-isolation -e .
 
 if %errorlevel% neq 0 exit /b %errorlevel%

ci/setup_env.sh

@@ -115,9 +115,20 @@ conda list pandas
 
 # Make sure any error below is reported as such
 
-echo "Build extensions and install pandas"
-python setup.py build_ext -q --inplace
-python -m pip install -e .
+echo "[Build extensions]"
+python setup.py build_ext -q -i
+
+# XXX: Some of our environments end up with old verisons of pip (10.x)
+# Adding a new enough verison of pip to the requirements explodes the
+# solve time. Just using pip to update itself.
+# - py35_macos
+# - py35_compat
+# - py36_32bit
+echo "[Updating pip]"
+python -m pip install --no-deps -U pip wheel setuptools
+
+echo "[Install pandas]"
+python -m pip install --no-build-isolation -e .
 
 echo
 echo "conda list"

doc/source/development/contributing.rst

@@ -208,7 +208,7 @@ We'll now kick off a three-step process:
 
    # Build and install pandas
    python setup.py build_ext --inplace -j 4
-   python -m pip install -e .
+   python -m pip install -e --no-build-isolation .
 
 At this point you should be able to import pandas from your locally built version::
 
@@ -252,7 +252,7 @@ You'll need to have at least python3.5 installed on your system.
 
    # Build and install pandas
    python setup.py build_ext --inplace -j 4
-   python -m pip install -e .
+   python -m pip install -e --no-build-isolation .
 
 Creating a branch
 -----------------

doc/source/whatsnew/v0.25.2.rst

@@ -52,7 +52,6 @@ Indexing
 -
 -
 -
--
 
 Missing
 ^^^^^^^

doc/source/whatsnew/v1.0.0.rst

@@ -170,6 +170,7 @@ Indexing
 
 - Bug in assignment using a reverse slicer (:issue:`26939`)
 - Bug in reindexing a :meth:`PeriodIndex` with another type of index that contained a `Period` (:issue:`28323`) (:issue:`28337`)
+- Fix assignment of column via `.loc` with numpy non-ns datetime type (:issue:`27395`)
 
 Missing
 ^^^^^^^
@@ -226,6 +227,7 @@ Sparse
 Build Changes
 ^^^^^^^^^^^^^
 - Fixed pyqt development dependency issue because of different pyqt package name in conda and PyPI (:issue:`26838`)
+- Added a `pyproject.toml <https://www.python.org/dev/peps/pep-0517/>`_ file (:issue:`20775`)
 
 
 ExtensionArray
@@ -240,7 +242,7 @@ Other
 - Trying to set the ``display.precision``, ``display.max_rows`` or ``display.max_columns`` using :meth:`set_option` to anything but a ``None`` or a positive int will raise a ``ValueError`` (:issue:`23348`)
 - Using :meth:`DataFrame.replace` with overlapping keys in a nested dictionary will no longer raise, now matching the behavior of a flat dictionary (:issue:`27660`)
 - :meth:`DataFrame.to_csv` and :meth:`Series.to_csv` now support dicts as ``compression`` argument with key ``'method'`` being the compression method and others as additional compression options when the compression method is ``'zip'``. (:issue:`26023`)
--
+- :meth:`Series.append` will no longer raise a ``TypeError`` when passed a tuple of ``Series`` (:issue:`28410`)
 
 .. _whatsnew_1000.contributors:
 

environment.yml

@@ -82,7 +82,3 @@ dependencies:
   - xlwt  # pandas.read_excel, DataFrame.to_excel, pandas.ExcelWriter, pandas.ExcelFile
   - odfpy  # pandas.read_excel
   - pyreadstat  # pandas.read_spss
-
-  # temporarily pin openssl
-  # https://github.com/pandas-dev/pandas/issues/28402
-  - openssl=1.1.1c

pandas/core/arrays/sparse.py

@@ -245,10 +245,10 @@ def construct_from_string(cls, string):
         if string.startswith("Sparse"):
             try:
                 sub_type, has_fill_value = cls._parse_subtype(string)
-                result = SparseDtype(sub_type)
-            except Exception:
+            except ValueError:
                 raise TypeError(msg)
             else:
+                result = SparseDtype(sub_type)
                 msg = (
                     "Could not construct SparseDtype from '{}'.\n\nIt "
                     "looks like the fill_value in the string is not "

pandas/core/dtypes/cast.py

@@ -1029,7 +1029,10 @@ def maybe_cast_to_datetime(value, dtype, errors="raise"):
             )
 
             if is_datetime64 and not is_dtype_equal(dtype, _NS_DTYPE):
-                if dtype.name in ("datetime64", "datetime64[ns]"):
+
+                # pandas supports dtype whose granularity is less than [ns]
+                # e.g., [ps], [fs], [as]
+                if dtype <= np.dtype("M8[ns]"):
                     if dtype.name == "datetime64":
                         raise ValueError(msg.format(dtype=dtype.name))
                     dtype = _NS_DTYPE
@@ -1047,7 +1050,10 @@ def maybe_cast_to_datetime(value, dtype, errors="raise"):
                     value = [value]
 
             elif is_timedelta64 and not is_dtype_equal(dtype, _TD_DTYPE):
-                if dtype.name in ("timedelta64", "timedelta64[ns]"):
+
+                # pandas supports dtype whose granularity is less than [ns]
+                # e.g., [ps], [fs], [as]
+                if dtype <= np.dtype("m8[ns]"):
                     if dtype.name == "timedelta64":
                         raise ValueError(msg.format(dtype=dtype.name))
                     dtype = _TD_DTYPE

pandas/core/dtypes/common.py

@@ -2051,10 +2051,8 @@ def pandas_dtype(dtype):
     # raise a consistent TypeError if failed
     try:
         npdtype = np.dtype(dtype)
-    except Exception:
-        # we don't want to force a repr of the non-string
-        if not isinstance(dtype, str):
-            raise TypeError("data type not understood")
+    except SyntaxError:
+        # np.dtype uses `eval` which can raise SyntaxError
         raise TypeError("data type '{}' not understood".format(dtype))
 
     # Any invalid dtype (such as pd.Timestamp) should raise an error.

pandas/core/ops/__init__.py

@@ -40,7 +40,11 @@
 
 from pandas._typing import ArrayLike
 from pandas.core.construction import array, extract_array
-from pandas.core.ops.array_ops import comp_method_OBJECT_ARRAY, define_na_arithmetic_op
+from pandas.core.ops.array_ops import (
+    comp_method_OBJECT_ARRAY,
+    define_na_arithmetic_op,
+    na_arithmetic_op,
+)
 from pandas.core.ops.docstrings import (
     _arith_doc_FRAME,
     _flex_comp_doc_FRAME,
@@ -627,12 +631,13 @@ def _arith_method_SERIES(cls, op, special):
         _construct_divmod_result if op in [divmod, rdivmod] else _construct_result
     )
 
-    na_op = define_na_arithmetic_op(op, str_rep, eval_kwargs)
-
     def wrapper(left, right):
         if isinstance(right, ABCDataFrame):
             return NotImplemented
 
+        left, right = _align_method_SERIES(left, right)
+        res_name = get_op_result_name(left, right)
+
         keep_null_freq = isinstance(
             right,
             (
@@ -644,9 +649,6 @@ def wrapper(left, right):
             ),
         )
 
-        left, right = _align_method_SERIES(left, right)
-        res_name = get_op_result_name(left, right)
-
         lvalues = extract_array(left, extract_numpy=True)
         rvalues = extract_array(right, extract_numpy=True)
 
@@ -659,7 +661,7 @@ def wrapper(left, right):
 
         else:
             with np.errstate(all="ignore"):
-                result = na_op(lvalues, rvalues)
+                result = na_arithmetic_op(lvalues, rvalues, op, str_rep, eval_kwargs)
 
         # We do not pass dtype to ensure that the Series constructor
         #  does inference in the case where `result` has object-dtype.

pandas/core/ops/array_ops.py

@@ -98,31 +98,37 @@ def masked_arith_op(x, y, op):
 
 def define_na_arithmetic_op(op, str_rep, eval_kwargs):
     def na_op(x, y):
-        """
-        Return the result of evaluating op on the passed in values.
+        return na_arithmetic_op(x, y, op, str_rep, eval_kwargs)
 
-        If native types are not compatible, try coersion to object dtype.
+    return na_op
 
-        Parameters
-        ----------
-        x : array-like
-        y : array-like or scalar
 
-        Returns
-        -------
-        array-like
+def na_arithmetic_op(left, right, op, str_rep, eval_kwargs):
+    """
+    Return the result of evaluating op on the passed in values.
 
-        Raises
-        ------
-        TypeError : invalid operation
-        """
-        import pandas.core.computation.expressions as expressions
+    If native types are not compatible, try coersion to object dtype.
 
-        try:
-            result = expressions.evaluate(op, str_rep, x, y, **eval_kwargs)
-        except TypeError:
-            result = masked_arith_op(x, y, op)
+    Parameters
+    ----------
+    left : np.ndarray
+    right : np.ndarray or scalar
+    str_rep : str or None
+    eval_kwargs : kwargs to pass to expressions
+
+    Returns
+    -------
+    array-like
+
+    Raises
+    ------
+    TypeError : invalid operation
+    """
+    import pandas.core.computation.expressions as expressions
 
-        return missing.dispatch_fill_zeros(op, x, y, result)
+    try:
+        result = expressions.evaluate(op, str_rep, left, right, **eval_kwargs)
+    except TypeError:
+        result = masked_arith_op(left, right, op)
 
-    return na_op
+    return missing.dispatch_fill_zeros(op, left, right, result)

pandas/core/series.py

@@ -2730,9 +2730,8 @@ def append(self, to_append, ignore_index=False, verify_integrity=False):
         from pandas.core.reshape.concat import concat
 
         if isinstance(to_append, (list, tuple)):
-            # https://github.com/pandas-dev/pandas/issues/28410
-            # error: Unsupported operand types for + ("List[Any]" and "Tuple[Any, ...]")
-            to_concat = [self] + to_append  # type: ignore
+            to_concat = [self]
+            to_concat.extend(to_append)
         else:
             to_concat = [self, to_append]
         return concat(

pandas/io/pytables.py

@@ -1791,7 +1791,7 @@ def convert(self, values, nan_rep, encoding, errors, start=None, stop=None):
         # making an Index instance could throw a number of different errors
         try:
             self.values = Index(values, **kwargs)
-        except Exception:  # noqa: E722
+        except Exception:
 
             # if the output freq is different that what we recorded,
             # it should be None (see also 'doc example part 2')

pandas/tests/dtypes/test_common.py

@@ -678,7 +678,7 @@ def test__get_dtype_sparse():
         (None, "Cannot deduce dtype from null object"),
         (1, "data type not understood"),
         (1.2, "data type not understood"),
-        ("random string", "data type 'random string' not understood"),
+        ("random string", 'data type "random string" not understood'),
         (pd.DataFrame([1, 2]), "data type not understood"),
     ],
 )

pandas/tests/indexes/interval/test_astype.py

@@ -67,7 +67,7 @@ def test_astype_cannot_cast(self, index, dtype):
             index.astype(dtype)
 
     def test_astype_invalid_dtype(self, index):
-        msg = "data type 'fake_dtype' not understood"
+        msg = 'data type "fake_dtype" not understood'
         with pytest.raises(TypeError, match=msg):
             index.astype("fake_dtype")
 

pandas/tests/indexes/interval/test_construction.py

@@ -164,7 +164,7 @@ def test_generic_errors(self, constructor):
             constructor(dtype="int64", **filler)
 
         # invalid dtype
-        msg = "data type 'invalid' not understood"
+        msg = 'data type "invalid" not understood'
         with pytest.raises(TypeError, match=msg):
             constructor(dtype="invalid", **filler)
 

pandas/tests/indexing/test_loc.py

@@ -690,6 +690,28 @@ def test_loc_setitem_consistency_slice_column_len(self):
         )
         tm.assert_series_equal(df[("Respondent", "Duration")], expected)
 
+    @pytest.mark.parametrize("unit", ["Y", "M", "D", "h", "m", "s", "ms", "us"])
+    def test_loc_assign_non_ns_datetime(self, unit):
+        # GH 27395, non-ns dtype assignment via .loc should work
+        # and return the same result when using simple assignment
+        df = DataFrame(
+            {
+                "timestamp": [
+                    np.datetime64("2017-02-11 12:41:29"),
+                    np.datetime64("1991-11-07 04:22:37"),
+                ]
+            }
+        )
+
+        df.loc[:, unit] = df.loc[:, "timestamp"].values.astype(
+            "datetime64[{unit}]".format(unit=unit)
+        )
+        df["expected"] = df.loc[:, "timestamp"].values.astype(
+            "datetime64[{unit}]".format(unit=unit)
+        )
+        expected = Series(df.loc[:, "expected"], name=unit)
+        tm.assert_series_equal(df.loc[:, unit], expected)
+
     def test_loc_setitem_frame(self):
         df = self.frame_labels
 

pandas/tests/io/parser/test_dtypes.py

@@ -79,7 +79,7 @@ def test_invalid_dtype_per_column(all_parsers):
 3,4.5
 4,5.5"""
 
-    with pytest.raises(TypeError, match="data type 'foo' not understood"):
+    with pytest.raises(TypeError, match='data type "foo" not understood'):
         parser.read_csv(StringIO(data), dtype={"one": "foo", 1: "int"})