Enable round in cudf for DataFrame and Series (#7022)

This enables round for DataFrames and Series using the libcudf round implementation and removes the old numba round implementation.

Closes #1270

Authors:
  - @ChrisJar

Approvers:
  - Ashwin Srinath (@shwina)
  - Michael Wang (@isVoid)
  - Ram (Ramakrishna Prabhu) (@rgsl888prabhu)
  - GALI PREM SAGAR (@galipremsagar)

URL: #7022

python/cudf/cudf/_lib/__init__.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2020, NVIDIA CORPORATION.
+# Copyright (c) 2020-2021, NVIDIA CORPORATION.
 import numpy as np
 
 from . import (
@@ -23,6 +23,7 @@
     replace,
     reshape,
     rolling,
+    round,
     search,
     sort,
     stream_compaction,

python/cudf/cudf/_lib/cpp/round.pxd

@@ -0,0 +1,19 @@
+# Copyright (c) 2021, NVIDIA CORPORATION.
+
+from libc.stdint cimport int32_t
+from libcpp.memory cimport unique_ptr
+
+from cudf._lib.cpp.column.column cimport column
+from cudf._lib.cpp.column.column_view cimport column_view
+
+cdef extern from "cudf/round.hpp" namespace "cudf" nogil:
+
+    ctypedef enum rounding_method "cudf::rounding_method":
+        HALF_UP "cudf::rounding_method::HALF_UP"
+        HALF_EVEN "cudf::rounding_method::HALF_EVEN"
+
+    cdef unique_ptr[column] round (
+        const column_view& input,
+        int32_t decimal_places,
+        rounding_method method,
+    ) except +

python/cudf/cudf/_lib/round.pyx

@@ -0,0 +1,42 @@
+# Copyright (c) 2021, NVIDIA CORPORATION.
+
+from libcpp.memory cimport unique_ptr
+from libcpp.utility cimport move
+
+from cudf._lib.column cimport Column
+
+from cudf._lib.cpp.column.column cimport column
+from cudf._lib.cpp.column.column_view cimport column_view
+from cudf._lib.cpp.round cimport (
+    rounding_method as cpp_rounding_method,
+    round as cpp_round
+)
+
+
+def round(Column input_col, int decimal_places=0):
+    """
+    Round column values to the given number of decimal places
+
+    Parameters
+    ----------
+    input_col : Column whose values will be rounded
+    decimal_places : The number or decimal places to round to
+
+    Returns
+    -------
+    A Column with values rounded to the given number of decimal places
+    """
+
+    cdef column_view input_col_view = input_col.view()
+    cdef unique_ptr[column] c_result
+
+    with nogil:
+        c_result = move(
+            cpp_round(
+                input_col_view,
+                decimal_places,
+                cpp_rounding_method.HALF_EVEN,
+            )
+        )
+
+    return Column.from_unique_ptr(move(c_result))

python/cudf/cudf/core/column/numerical.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2020, NVIDIA CORPORATION.
+# Copyright (c) 2018-2021, NVIDIA CORPORATION.
 
 from numbers import Number
 
@@ -342,17 +342,9 @@ def corr(self, other):
         return cov / lhs_std / rhs_std
 
     def round(self, decimals=0):
-        if decimals < 0:
-            msg = "Decimal values < 0 are not yet supported."
-            raise NotImplementedError(msg)
-
-        if np.issubdtype(self.dtype, np.integer):
-            return self
-
-        data = Buffer(
-            cudautils.apply_round(self.data_array_view, decimals).view("|u1")
-        )
-        return column.build_column(data=data, dtype=self.dtype, mask=self.mask)
+        """Round the values in the Column to the given number of decimals.
+        """
+        return libcudf.round.round(self, decimal_places=decimals)
 
     def applymap(self, udf, out_dtype=None):
         """Apply an element-wise function to transform the values in the Column.

python/cudf/cudf/core/frame.py

@@ -1751,6 +1751,117 @@ def __arrow_array__(self, type=None):
             "consider using .to_arrow()"
         )
 
+    def round(self, decimals=0):
+        """
+        Round a DataFrame to a variable number of decimal places.
+
+        Parameters
+        ----------
+        decimals : int, dict, Series
+            Number of decimal places to round each column to. If an int is
+            given, round each column to the same number of places.
+            Otherwise dict and Series round to variable numbers of places.
+            Column names should be in the keys if `decimals` is a
+            dict-like, or in the index if `decimals` is a Series. Any
+            columns not included in `decimals` will be left as is. Elements
+            of `decimals` which are not columns of the input will be
+            ignored.
+
+        Returns
+        -------
+        DataFrame
+            A DataFrame with the affected columns rounded to the specified
+            number of decimal places.
+
+        Examples
+        --------
+        >>> df = cudf.DataFrame(
+                [(.21, .32), (.01, .67), (.66, .03), (.21, .18)],
+        ...     columns=['dogs', 'cats']
+        ... )
+        >>> df
+            dogs  cats
+        0  0.21  0.32
+        1  0.01  0.67
+        2  0.66  0.03
+        3  0.21  0.18
+
+        By providing an integer each column is rounded to the same number
+        of decimal places
+
+        >>> df.round(1)
+            dogs  cats
+        0   0.2   0.3
+        1   0.0   0.7
+        2   0.7   0.0
+        3   0.2   0.2
+
+        With a dict, the number of places for specific columns can be
+        specified with the column names as key and the number of decimal
+        places as value
+
+        >>> df.round({'dogs': 1, 'cats': 0})
+            dogs  cats
+        0   0.2   0.0
+        1   0.0   1.0
+        2   0.7   0.0
+        3   0.2   0.0
+
+        Using a Series, the number of places for specific columns can be
+        specified with the column names as index and the number of
+        decimal places as value
+
+        >>> decimals = cudf.Series([0, 1], index=['cats', 'dogs'])
+        >>> df.round(decimals)
+            dogs  cats
+        0   0.2   0.0
+        1   0.0   1.0
+        2   0.7   0.0
+        3   0.2   0.0
+        """
+
+        if isinstance(decimals, cudf.Series):
+            decimals = decimals.to_pandas()
+
+        if isinstance(decimals, (dict, pd.Series)):
+            if (
+                isinstance(decimals, pd.Series)
+                and not decimals.index.is_unique
+            ):
+                raise ValueError("Index of decimals must be unique")
+
+            cols = {
+                name: col.round(decimals[name])
+                if (
+                    name in decimals.keys()
+                    and pd.api.types.is_numeric_dtype(col.dtype)
+                )
+                else col.copy(deep=True)
+                for name, col in self._data.items()
+            }
+        elif isinstance(decimals, int):
+            cols = {
+                name: col.round(decimals)
+                if pd.api.types.is_numeric_dtype(col.dtype)
+                else col.copy(deep=True)
+                for name, col in self._data.items()
+            }
+        else:
+            raise TypeError(
+                "decimals must be an integer, a dict-like or a Series"
+            )
+
+        return self.__class__._from_table(
+            Frame(
+                data=cudf.core.column_accessor.ColumnAccessor(
+                    cols,
+                    multiindex=self._data.multiindex,
+                    level_names=self._data.level_names,
+                )
+            ),
+            index=self._index,
+        )
+
     @annotate("SAMPLE", color="orange", domain="cudf_python")
     def sample(
         self,

python/cudf/cudf/core/series.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2020, NVIDIA CORPORATION.
+# Copyright (c) 2018-2021, NVIDIA CORPORATION.
 import pickle
 import warnings
 from collections import abc as abc
@@ -3506,8 +3506,31 @@ def mode(self, dropna=True):
         return Series(val_counts.index.sort_values(), name=self.name)
 
     def round(self, decimals=0):
-        """Round a Series to a configurable number of decimal places.
         """
+        Round each value in a Series to the given number of decimals.
+
+        Parameters
+        ----------
+        decimals : int, default 0
+            Number of decimal places to round to. If decimals is negative,
+            it specifies the number of positions to the left of the decimal
+            point.
+
+        Returns
+        -------
+        Series
+            Rounded values of the Series.
+
+        Examples
+        --------
+        >>> s = cudf.Series([0.1, 1.4, 2.9])
+        >>> s.round()
+        0    0.0
+        1    1.0
+        2    3.0
+        dtype: float64
+        """
+
         return Series(
             self._column.round(decimals=decimals),
             name=self.name,

python/cudf/cudf/tests/test_dataframe.py

@@ -3243,86 +3243,61 @@ def test_ndim():
 
 
 @pytest.mark.parametrize(
-    "arr",
-    [
-        np.random.normal(-100, 100, 1000),
-        np.random.randint(-50, 50, 1000),
-        np.zeros(100),
-        np.repeat([-0.6459412758761901], 100),
-        np.repeat(np.nan, 100),
-        np.array([1.123, 2.343, np.nan, 0.0]),
-    ],
-)
-@pytest.mark.parametrize(
-    "decimal",
+    "decimals",
     [
+        -3,
         0,
-        1,
-        2,
-        3,
-        4,
         5,
-        6,
-        7,
-        8,
-        9,
-        10,
-        11,
-        12,
-        13,
-        14,
-        15,
-        16,
-        17,
-        pytest.param(
-            -1,
-            marks=[
-                pytest.mark.xfail(reason="NotImplementedError: decimals < 0")
-            ],
-        ),
+        pd.Series([1, 4, 3, -6], index=["w", "x", "y", "z"]),
+        gd.Series([-4, -2, 12], index=["x", "y", "z"]),
+        {"w": -1, "x": 15, "y": 2},
     ],
 )
-def test_round(arr, decimal):
-    pser = pd.Series(arr)
-    ser = gd.Series(arr)
-    result = ser.round(decimal)
-    expected = pser.round(decimal)
+def test_dataframe_round(decimals):
+    pdf = pd.DataFrame(
+        {
+            "w": np.arange(0.5, 10.5, 1),
+            "x": np.random.normal(-100, 100, 10),
+            "y": np.array(
+                [
+                    14.123,
+                    2.343,
+                    np.nan,
+                    0.0,
+                    -8.302,
+                    np.nan,
+                    94.313,
+                    -112.236,
+                    -8.029,
+                    np.nan,
+                ]
+            ),
+            "z": np.repeat([-0.6459412758761901], 10),
+        }
+    )
+    gdf = gd.DataFrame.from_pandas(pdf)
 
+    if isinstance(decimals, gd.Series):
+        pdecimals = decimals.to_pandas()
+    else:
+        pdecimals = decimals
+
+    result = gdf.round(decimals)
+    expected = pdf.round(pdecimals)
     assert_eq(result, expected)
 
     # with nulls, maintaining existing null mask
-    arr = arr.astype("float64")  # for pandas nulls
-    mask = np.random.randint(0, 2, arr.shape[0])
-    arr[mask == 1] = np.nan
+    for c in pdf.columns:
+        arr = pdf[c].to_numpy().astype("float64")  # for pandas nulls
+        arr.ravel()[np.random.choice(10, 5, replace=False)] = np.nan
+        pdf[c] = gdf[c] = arr
 
-    pser = pd.Series(arr)
-    ser = gd.Series(arr)
-    result = ser.round(decimal)
-    expected = pser.round(decimal)
+    result = gdf.round(decimals)
+    expected = pdf.round(pdecimals)
 
     assert_eq(result, expected)
-    np.array_equal(ser.nullmask.to_array(), result.to_array())
-
-
-@pytest.mark.parametrize(
-    "series",
-    [
-        gd.Series([1.0, None, np.nan, 4.0], nan_as_null=False),
-        gd.Series([1.24430, None, np.nan, 4.423530], nan_as_null=False),
-        gd.Series([1.24430, np.nan, 4.423530], nan_as_null=False),
-        gd.Series([-1.24430, np.nan, -4.423530], nan_as_null=False),
-        gd.Series(np.repeat(np.nan, 100)),
-    ],
-)
-@pytest.mark.parametrize("decimal", [0, 1, 2, 3])
-def test_round_nan_as_null_false(series, decimal):
-    pser = series.to_pandas()
-    ser = gd.Series(series)
-    result = ser.round(decimal)
-    expected = pser.round(decimal)
-    np.testing.assert_array_almost_equal(
-        result.to_pandas(), expected, decimal=10
-    )
+    for c in gdf.columns:
+        np.array_equal(gdf[c].nullmask.to_array(), result[c].to_array())
 
 
 @pytest.mark.parametrize(