[API Compatibilities] Add rand_like, multinomial and var (#74920)

* add rand_like, multinomial and var

* fix rand_like impl

* update rand_like, rand, uniform implementation

* aligning logic with rand_like

* remove redundant operations

Author: LLSGYN

python/paddle/__init__.py

@@ -661,6 +661,7 @@ def new_init(self, *args, **kwargs):
     normal_,
     poisson,
     rand,
+    rand_like,
     randint,
     randint_like,
     randn,
@@ -1252,6 +1253,7 @@ def __dir__(self):
     'geometric_',
     'randn',
     'randn_like',
+    'rand_like',
     'strided_slice',
     'unique',
     'unique_consecutive',

python/paddle/tensor/__init__.py

@@ -454,6 +454,7 @@
     normal_,
     poisson,
     rand,
+    rand_like,
     randint,
     randint_like,
     randn,

python/paddle/tensor/random.py

@@ -453,6 +453,8 @@ def multinomial(
     num_samples: int = 1,
     replacement: bool = False,
     name: str | None = None,
+    *,
+    out: Tensor | None = None,
 ) -> Tensor:
     """
     Returns a Tensor filled with random values sampled from a Multinomial
@@ -474,6 +476,7 @@ def multinomial(
         name(str|None, optional): The default value is None. Normally there is no
             need for user to set this property. For more information, please
             refer to :ref:`api_guide_Name`.
+        out (Tensor|None, optional): The output Tensor. If set, the result will be stored in this Tensor. Default is None.
     Returns:
         Tensor, A Tensor filled with sampled category index after ``num_samples`` times samples.
 
@@ -516,7 +519,7 @@ def multinomial(
     """
 
     if in_dynamic_or_pir_mode():
-        return _C_ops.multinomial(x, num_samples, replacement)
+        return _C_ops.multinomial(x, num_samples, replacement, out=out)
     else:
         check_variable_and_dtype(
             x, "x", ["uint16", "float16", "float32", "float64"], "multinomial"
@@ -1150,14 +1153,104 @@ def randn_like(
     """
     if dtype is None:
         dtype = x.dtype
-    else:
-        if not isinstance(dtype, (core.VarDesc.VarType, core.DataType)):
-            dtype = convert_np_dtype_to_dtype_(dtype)
     shape = paddle.shape(x)
 
     return standard_normal(shape, dtype, name)
 
 
+def rand_like(
+    input,
+    name: str | None = None,
+    *,
+    dtype: DTypeLike | None = None,
+    device: PlaceLike | None = None,
+    requires_grad: bool = False,
+):
+    """
+    Returns a tensor with the same size as input that is filled with random numbers from a uniform distribution on the interval [0, 1).
+
+    Args:
+        input (Tensor): The input multi-dimensional tensor which specifies shape. The dtype of ``input``
+            can be float16, float64, float8_e4m3fn, float32, bfloat16.
+        name (str|None, optional): The default value is None. Normally there is no
+            need for user to set this property. For more information, please
+            refer to :ref:`api_guide_Name`.
+        dtype (str|np.dtype|paddle.dtype|None, optional): The data type of the
+            output tensor. Supported data types: float16, float64, float8_e4m3fn, float32, bfloat16.
+            If ``dtype`` is None, the data type is the same as input's data type. Default is None.
+        device (str|paddle.Place|None, optional): The device on which to place the created tensor.
+            If None, the device is the same as input's device. Default is None.
+        requires_grad (bool, optional): Whether to compute gradients for the created tensor.
+            Default is False.
+
+    Returns:
+        Tensor: A Tensor with the same size as input that is filled with random numbers from a uniform distribution on the interval [0, 1).
+
+    Examples:
+        .. code-block:: python
+
+            >>> import paddle
+
+            >>> # example 1:
+            >>> # dtype is None and the dtype of input is float32
+            >>> x = paddle.zeros((2, 3)).astype("float32")
+            >>> out1 = paddle.rand_like(x)
+            >>> print(out1)
+            >>> # doctest: +SKIP("Random output")
+            Tensor(shape=[2, 3], dtype=float32, place=Place(cpu), stop_gradient=True,
+            [[0.34962332, 0.82356787, 0.91275704],
+             [0.12328923, 0.58439839, 0.32735515]])
+            >>> # doctest: -SKIP
+            >>> print(out1.dtype)
+            paddle.float32
+
+            >>> # example 2:
+            >>> # dtype is None and the dtype of input is float64
+            >>> x = paddle.zeros((2, 3)).astype("float64")
+            >>> out2 = paddle.rand_like(x)
+            >>> print(out2)
+            >>> # doctest: +SKIP("Random output")
+            Tensor(shape=[2, 3], dtype=float64, place=Place(cpu), stop_gradient=True,
+            [[0.73964721, 0.28413662, 0.91918457],
+             [0.62838351, 0.39185921, 0.51561823]])
+            >>> # doctest: -SKIP
+            >>> print(out2.dtype)
+            paddle.float64
+
+            >>> # example 3:
+            >>> # dtype is float64 and the dtype of input is float32
+            >>> x = paddle.zeros((2, 3)).astype("float32")
+            >>> out3 = paddle.rand_like(x, dtype="float64")
+            >>> print(out3)
+            >>> # doctest: +SKIP("Random output")
+            Tensor(shape=[2, 3], dtype=float64, place=Place(cpu), stop_gradient=True,
+            [[0.84492219, 0.11572551, 0.73868765],
+             [0.90269387, 0.45644298, 0.28739912]])
+            >>> # doctest: -SKIP
+            >>> print(out3.dtype)
+            paddle.float64
+
+            >>> # example 4:
+            >>> # with requires_grad=True
+            >>> x = paddle.zeros((2, 2)).astype("float32")
+            >>> out4 = paddle.rand_like(x, requires_grad=True)
+            >>> print(out4.stop_gradient)
+            False
+    """
+    if dtype is None:
+        dtype = input.dtype
+
+    return uniform(
+        shape=input.shape,
+        dtype=dtype,
+        min=0.0,
+        max=1.0,
+        name=name,
+        device=device,
+        requires_grad=requires_grad,
+    )
+
+
 def normal(
     mean: complex | Tensor = 0.0,
     std: float | Tensor = 1.0,
@@ -1370,6 +1463,10 @@ def uniform(
     max: float = 1.0,
     seed: int = 0,
     name: str | None = None,
+    *,
+    out: Tensor | None = None,
+    device: PlaceLike | None = None,
+    requires_grad: bool = False,
 ) -> Tensor:
     """
     Returns a Tensor filled with random values sampled from a uniform
@@ -1460,14 +1557,23 @@ def uniform(
 
     if in_dynamic_mode():
         shape = paddle.utils.convert_shape_to_list(shape)
-        return _C_ops.uniform(
+        place = (
+            _current_expected_place()
+            if device is None
+            else _get_paddle_place(device)
+        )
+        tensor = _C_ops.uniform(
             shape,
             dtype,
             float(min),
             float(max),
             seed,
-            _current_expected_place(),
+            place,
+            out=out,
         )
+        if requires_grad is True:
+            tensor.stop_gradient = False
+        return tensor
     elif in_pir_mode():
         check_type(
             shape, 'shape', (list, tuple, paddle.pir.Value), 'uniform/rand'
@@ -1482,14 +1588,23 @@ def uniform(
         if isinstance(max, int):
             max = float(max)
 
-        return _C_ops.uniform(
+        place = (
+            _current_expected_place()
+            if device is None
+            else _get_paddle_place(device)
+        )
+        tensor = _C_ops.uniform(
             shape,
             dtype,
             min,
             max,
             seed,
-            _current_expected_place(),
+            place,
+            out=out,
         )
+        if requires_grad is True:
+            tensor.stop_gradient = False
+        return tensor
     else:
         check_type(shape, 'shape', (list, tuple, Variable), 'uniform/rand')
         check_dtype(dtype, 'dtype', supported_dtypes, 'uniform/rand')

python/paddle/tensor/stat.py

@@ -32,7 +32,10 @@
 
 from ..base.data_feeder import check_type, check_variable_and_dtype
 from ..common_ops_import import Variable
-from ..framework import LayerHelper, core
+from ..framework import (
+    LayerHelper,
+    core,
+)
 from .math import _get_reduce_axis_with_tensor
 
 if TYPE_CHECKING:
@@ -157,9 +160,12 @@ def mean(
 def var(
     x: Tensor,
     axis: int | Sequence[int] | None = None,
-    unbiased: bool = True,
+    unbiased: bool | None = None,
     keepdim: bool = False,
     name: str | None = None,
+    *,
+    correction: float = 1,
+    out: Tensor | None = None,
 ) -> Tensor:
     """
     Computes the variance of ``x`` along ``axis`` .
@@ -181,6 +187,9 @@ def var(
         unbiased (bool, optional): Whether to use the unbiased estimation. If ``unbiased`` is True, the divisor used in the computation is :math:`N - 1`, where :math:`N` represents the number of elements along ``axis`` , otherwise the divisor is :math:`N`. Default is True.
         keep_dim (bool, optional): Whether to reserve the reduced dimension in the output Tensor. The result tensor will have one fewer dimension than the input unless keep_dim is true. Default is False.
         name (str|None, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
+        correction (int|float, optional): Difference between the sample size and sample degrees of freedom.
+            Defaults to 1 (Bessel's correction). If unbiased is specified, this parameter is ignored.
+        out (Tensor|None, optional): Output tensor. Default is None.
 
     Returns:
         Tensor, results of variance along ``axis`` of ``x``, with the same data type as ``x``.
@@ -198,28 +207,41 @@ def var(
             >>> print(out2.numpy())
             [1.         4.3333335]
     """
+    if unbiased is not None and correction != 1:
+        raise ValueError("Only one of unbiased and correction may be given")
+
+    if unbiased is not None:
+        actual_correction = 1.0 if unbiased else 0.0
+    else:
+        actual_correction = float(correction)
     if not in_dynamic_mode():
         check_variable_and_dtype(
             x, 'x', ['float16', 'float32', 'float64'], 'var'
         )
 
     u = mean(x, axis, True, name)
     dtype = paddle.float32 if x.dtype == paddle.float16 else x.dtype
-    out = paddle.sum(
+    out_tensor = paddle.sum(
         paddle.pow((x - u), 2), axis, keepdim=keepdim, name=name, dtype=dtype
     )
 
     n = paddle.cast(paddle.numel(x), "int64") / paddle.cast(
-        paddle.numel(out), "int64"
+        paddle.numel(out_tensor), "int64"
     )
     n = n.astype(dtype)
-    if unbiased:
-        one_const = paddle.ones([], x.dtype)
-        if paddle.in_dynamic_mode() and n <= one_const:
+
+    if actual_correction != 0:
+        corrected_n = n - actual_correction
+        corrected_n = paddle.maximum(
+            corrected_n, paddle.zeros_like(corrected_n)
+        )
+        if paddle.in_dynamic_mode() and paddle.any(corrected_n <= 0):
             warnings.warn("Degrees of freedom is <= 0.", stacklevel=2)
-        n = n - 1.0
-    n.stop_gradient = True
-    out /= n
+    else:
+        corrected_n = n
+
+    corrected_n.stop_gradient = True
+    out_tensor /= corrected_n
 
     def _replace_nan(out):
         indices = paddle.arange(out.numel(), dtype='int64')
@@ -229,12 +251,20 @@ def _replace_nan(out):
         return out_nan
 
     if 0 in x.shape:
-        out = _replace_nan(out)
-    if len(x.shape) == 0 and not unbiased:
-        out = paddle.to_tensor(0, stop_gradient=out.stop_gradient)
-    if out.dtype != x.dtype:
-        return out.astype(x.dtype)
-    return out
+        out_tensor = _replace_nan(out_tensor)
+    if len(x.shape) == 0 and actual_correction == 0:
+        out_tensor = paddle.to_tensor(0, stop_gradient=out_tensor.stop_gradient)
+
+    if out_tensor.dtype != x.dtype:
+        result = out_tensor.astype(x.dtype)
+    else:
+        result = out_tensor
+
+    if out is not None:
+        paddle.assign(result, out)
+        return out
+
+    return result
 
 
 def std(