Add the op of unique_with_counts, expand count function of the op unique (#18720)
* test=develop Add the op of unique_with_counts, the op is calc the unqiue input of data, and output the corresponding indices and count of data. * test=develop Check the input and dtype in the op of unique_with_counts * test=develop test=document_preview update the API.spec for `unique_with_counts`, at the same time, optimize the python api in the op of `unique_with_count` * test=develop test=document_preview Fix some python api problem in the op of `unique_with_counts`, and change the error messsage in this op. * Fix some API problem in the op of `unique_with_counts` test=develop test=document_preview * test=develop test=document_preview Fix the api sample of op `unique_with_counts`, and update api.specpadding_in_crf
wawltor
6 years ago
committed by
GitHub
parent
5cf2d38594
commit
3ab1866ca5
@ -0,0 +1,71 @@
|
||||
/* Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License. */
|
||||
|
||||
#include "paddle/fluid/operators/unique_with_counts_op.h"
|
||||
|
||||
namespace paddle {
|
||||
namespace operators {
|
||||
|
||||
class UniqueWithCountsOp : public framework::OperatorWithKernel {
|
||||
public:
|
||||
using framework::OperatorWithKernel::OperatorWithKernel;
|
||||
|
||||
void InferShape(framework::InferShapeContext* ctx) const override {
|
||||
PADDLE_ENFORCE(ctx->HasInput("X"),
|
||||
"Input(X) of UniqueWithCountsOp should not be null.");
|
||||
PADDLE_ENFORCE(ctx->HasOutput("Out"),
|
||||
"Output(Out) of UniqueWithCountsOp should not be null.");
|
||||
PADDLE_ENFORCE(ctx->HasOutput("Index"),
|
||||
"Output(Index) of UniqueWithCountsOp should not be null.");
|
||||
PADDLE_ENFORCE(ctx->HasOutput("Count"),
|
||||
"Output(Count) of UniqueWithCountsOp should not be null.");
|
||||
|
||||
auto in_dims = ctx->GetInputDim("X");
|
||||
PADDLE_ENFORCE(in_dims.size() == 1,
|
||||
"The op of fluid.layers.unique_with_counts, Input(X) should "
|
||||
"be a vector.");
|
||||
|
||||
ctx->SetOutputDim("Out", {-1});
|
||||
ctx->SetOutputDim("Index", in_dims);
|
||||
ctx->SetOutputDim("Count", {-1});
|
||||
}
|
||||
};
|
||||
|
||||
class UniqueWithCountsOpMaker : public framework::OpProtoAndCheckerMaker {
|
||||
public:
|
||||
void Make() override {
|
||||
AddInput("X", "Input tensor. It should be a 1-D tensor.");
|
||||
AddAttr<int>("dtype", "data type for output index");
|
||||
AddOutput("Out", "A unique subsequence for input tensor.");
|
||||
AddOutput("Index",
|
||||
"An index tensor pointing to unique subsequence, which has "
|
||||
"identical shape with input tensor and the data type is set by "
|
||||
"the attr `dtype`");
|
||||
AddOutput("Count", "A subsequence for the count of unique index");
|
||||
AddComment(R"DOC(
|
||||
Return a unique subsequence for 1-D input tensor, index tensor pointing to this unique subsequence,
|
||||
and the subsequence for the count of unique index.
|
||||
)DOC");
|
||||
}
|
||||
};
|
||||
} // namespace operators
|
||||
} // namespace paddle
|
||||
|
||||
namespace ops = paddle::operators;
|
||||
REGISTER_OP_WITHOUT_GRADIENT(unique_with_counts, ops::UniqueWithCountsOp,
|
||||
ops::UniqueWithCountsOpMaker);
|
||||
REGISTER_OP_CPU_KERNEL(unique_with_counts, ops::UniqueWithCountsKernel<float>,
|
||||
ops::UniqueWithCountsKernel<double>,
|
||||
ops::UniqueWithCountsKernel<int32_t>,
|
||||
ops::UniqueWithCountsKernel<int64_t>);
|
||||
@ -0,0 +1,43 @@
|
||||
/* Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License. */
|
||||
|
||||
#pragma once
|
||||
#include <cmath>
|
||||
#include <unordered_map>
|
||||
#include <utility>
|
||||
#include <vector>
|
||||
#include "paddle/fluid/framework/op_registry.h"
|
||||
#include "paddle/fluid/operators/math/math_function.h"
|
||||
#include "paddle/fluid/operators/unique_op.h"
|
||||
|
||||
namespace paddle {
|
||||
namespace operators {
|
||||
|
||||
template <typename T>
|
||||
class UniqueWithCountsKernel : public framework::OpKernel<T> {
|
||||
public:
|
||||
void Compute(const framework::ExecutionContext& context) const override {
|
||||
auto data_type = static_cast<framework::proto::VarType::Type>(
|
||||
context.Attr<int>("dtype"));
|
||||
auto* x = context.Input<framework::Tensor>("X");
|
||||
auto* out = context.Output<framework::Tensor>("Out");
|
||||
auto* index = context.Output<framework::Tensor>("Index");
|
||||
auto* count = context.Output<framework::Tensor>("Count");
|
||||
framework::VisitDataType(data_type,
|
||||
UniqueOpFunctor<T>(out, index, x, count));
|
||||
}
|
||||
};
|
||||
|
||||
} // namespace operators
|
||||
} // namespace paddle
|
||||
@ -0,0 +1,84 @@
|
||||
# Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from __future__ import print_function
|
||||
|
||||
import unittest
|
||||
import numpy as np
|
||||
from op_test import OpTest
|
||||
import paddle.fluid.core as core
|
||||
from paddle.fluid.op import Operator
|
||||
|
||||
|
||||
class TestUniqueWithCountsOp(OpTest):
|
||||
def setUp(self):
|
||||
self.op_type = "unique_with_counts"
|
||||
self.init_config()
|
||||
|
||||
def test_check_output(self):
|
||||
self.check_output()
|
||||
|
||||
def init_config(self):
|
||||
self.inputs = {'X': np.array([2, 3, 3, 1, 5, 3], dtype='int64'), }
|
||||
self.attrs = {'dtype': int(core.VarDesc.VarType.INT32)}
|
||||
self.outputs = {
|
||||
'Out': np.array(
|
||||
[2, 3, 1, 5], dtype='int64'),
|
||||
'Index': np.array(
|
||||
[0, 1, 1, 2, 3, 1], dtype='int32'),
|
||||
'Count': np.array(
|
||||
[1, 3, 1, 1], dtype='int32')
|
||||
}
|
||||
|
||||
|
||||
class TestOne(TestUniqueWithCountsOp):
|
||||
def init_config(self):
|
||||
self.inputs = {'X': np.array([2], dtype='int64'), }
|
||||
self.attrs = {'dtype': int(core.VarDesc.VarType.INT32)}
|
||||
self.outputs = {
|
||||
'Out': np.array(
|
||||
[2], dtype='int64'),
|
||||
'Index': np.array(
|
||||
[0], dtype='int32'),
|
||||
'Count': np.array(
|
||||
[1], dtype='int32')
|
||||
}
|
||||
|
||||
|
||||
class TestRandom(TestUniqueWithCountsOp):
|
||||
def init_config(self):
|
||||
input_data = np.random.randint(0, 100, (2000, ), dtype='int64')
|
||||
self.inputs = {'X': input_data}
|
||||
self.attrs = {'dtype': int(core.VarDesc.VarType.INT64)}
|
||||
np_unique, np_index, reverse_index = np.unique(self.inputs['X'], True,
|
||||
True)
|
||||
np_tuple = [(np_unique[i], np_index[i]) for i in range(len(np_unique))]
|
||||
np_tuple.sort(key=lambda x: x[1])
|
||||
target_out = np.array([i[0] for i in np_tuple], dtype='int64')
|
||||
target_index = np.array(
|
||||
[list(target_out).index(i) for i in self.inputs['X']],
|
||||
dtype='int64')
|
||||
count = [0 for i in range(len(np_unique))]
|
||||
for i in range(target_index.shape[0]):
|
||||
count[target_index[i]] += 1
|
||||
target_count = np.array(count, dtype='int64')
|
||||
self.outputs = {
|
||||
'Out': target_out,
|
||||
'Index': target_index,
|
||||
'Count': target_count
|
||||
}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
unittest.main()
|
||||
Loading…
Reference in new issue