gx

chenyc

2025-02-12 ea42ff3ebee1eeb3fb29423aa848a249441db81c

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
/**
 * @license
 * Copyright 2020 Google LLC. 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.
 * =============================================================================
 */
import { ENGINE } from '../engine';
import { Diag } from '../kernel_names';
import { convertToTensor } from '../tensor_util_env';
import { op } from './operation';
/**
 * Returns a diagonal tensor with given diagonal values.
 *
 * Given a diagonal, this operation returns a tensor with the diagonal and
 * everything else padded with zeros.
 *
 * Assume the input has dimensions `[D1,..., Dk]`, then the output is a tensor
 * of rank 2k with dimensions `[D1,..., Dk, D1,..., Dk]`
 *
 * ```js
 * const x = tf.tensor1d([1, 2, 3, 4]);
 *
 * tf.diag(x).print()
 * ```
 * ```js
 * const x = tf.tensor2d([1, 2, 3, 4, 5, 6, 7, 8], [4, 2])
 *
 * tf.diag(x).print()
 * ```
 * @param x The input tensor.
 *
 * @doc {heading: 'Tensors', subheading: 'Creation'}
 */
function diag_(x) {
    const $x = convertToTensor(x, 'x', 'diag');
    const inputs = { x: $x };
    return ENGINE.runKernel(Diag, inputs);
}
export const diag = /* @__PURE__ */ op({ diag_ });
//# sourceMappingURL=data:application/json;base64,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