@tensorflow/tfjs-core
Version:
Hardware-accelerated JavaScript library for machine intelligence
154 lines • 6.46 kB
JavaScript
/**
* @license
* Copyright 2020 Google Inc. 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 * as tf from '../index';
import { ALL_ENVS, describeWithFlags } from '../jasmine_util';
import { expectArraysClose, expectArraysEqual } from '../test_util';
import * as reduce_util from './reduce_util';
describeWithFlags('argmax', ALL_ENVS, () => {
it('Tensor1D', async () => {
const a = tf.tensor1d([1, 0, 3, 2]);
const result = tf.argMax(a);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), 2);
});
it('one value', async () => {
const a = tf.tensor1d([10]);
const result = tf.argMax(a);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), 0);
});
it('N > than parallelization threshold', async () => {
const n = reduce_util.PARALLELIZE_THRESHOLD * 2;
const values = new Float32Array(n);
for (let i = 0; i < n; i++) {
values[i] = i;
}
const a = tf.tensor1d(values);
const result = tf.argMax(a);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), n - 1);
});
it('3D, N > than parallelization threshold', async () => {
const n = reduce_util.PARALLELIZE_THRESHOLD * 2;
const values = new Float32Array(n);
for (let i = 0; i < n; i++) {
values[i] = i;
}
const a = tf.tensor3d(values, [1, 1, n]);
const result = tf.argMax(a, -1);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), n - 1);
});
it('max index corresponds to start of a non-initial window', async () => {
const n = reduce_util.PARALLELIZE_THRESHOLD * 2;
const windowSize = reduce_util.computeOptimalWindowSize(n);
const values = new Float32Array(n);
const index = windowSize * 2;
values[index] = 1;
const a = tf.tensor1d(values);
const result = tf.argMax(a);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), index);
});
it('5D, max index corresponds to start of a non-initial window', async () => {
const n = reduce_util.PARALLELIZE_THRESHOLD * 2;
const windowSize = reduce_util.computeOptimalWindowSize(n);
const values = new Float32Array(n);
const index = windowSize * 2;
values[index] = 1;
const a = tf.tensor5d(values, [1, 1, 1, 1, n]);
const result = tf.argMax(a, -1);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), index);
});
it('ignores NaNs', async () => {
const a = tf.tensor1d([0, 3, 5, NaN, 3]);
const res = tf.argMax(a);
expect(res.dtype).toBe('int32');
expectArraysEqual(await res.data(), 2);
});
it('2D, no axis specified', async () => {
const a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
expectArraysEqual(await tf.argMax(a).data(), [1, 0, 1]);
});
it('4D, no axis specified', async () => {
const a = tf.tensor4d([3, -1, 0, 100, -7, 2], [2, 1, 1, 3]);
expectArraysEqual(await tf.argMax(a).data(), [1, 0, 1]);
});
it('2D, axis=0', async () => {
const a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
const r = tf.argMax(a, 0);
expect(r.shape).toEqual([3]);
expect(r.dtype).toBe('int32');
expectArraysEqual(await r.data(), [1, 0, 1]);
});
it('6D, axis=0', async () => {
const a = tf.tensor6d([3, -1, 0, 100, -7, 2], [2, 1, 1, 1, 1, 3]);
const r = tf.argMax(a, 0);
expect(r.shape).toEqual([1, 1, 1, 1, 3]);
expect(r.dtype).toBe('int32');
expectArraysEqual(await r.data(), [1, 0, 1]);
});
it('2D, axis=1', async () => {
const a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
const r = tf.argMax(a, 1);
expect(r.dtype).toBe('int32');
expectArraysEqual(await r.data(), [2, 0]);
});
it('2D, axis = -1', async () => {
const a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
const r = tf.argMax(a, -1);
expect(r.dtype).toBe('int32');
expectArraysEqual(await r.data(), [2, 0]);
});
it('throws when passed a non-tensor', () => {
expect(() => tf.argMax({}))
.toThrowError(/Argument 'x' passed to 'argMax' must be a Tensor/);
});
it('accepts a tensor-like object', async () => {
const result = tf.argMax([1, 0, 3, 2]);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), 2);
});
it('accepts tensor with bool values', async () => {
const t = tf.tensor1d([0, 1], 'bool');
const result = tf.argMax(t);
expect(result.dtype).toBe('int32');
expectArraysEqual(await result.data(), 1);
});
it('has gradient', async () => {
const a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
const dy = tf.ones([3], 'float32');
const da = tf.grad((x) => tf.argMax(x))(a, dy);
expect(da.dtype).toBe('float32');
expect(da.shape).toEqual([2, 3]);
expectArraysClose(await da.data(), [0, 0, 0, 0, 0, 0]);
});
it('gradient with clones', async () => {
const a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
const dy = tf.ones([3], 'float32');
const da = tf.grad((x) => tf.argMax(x.clone()).clone())(a, dy);
expect(da.dtype).toBe('float32');
expect(da.shape).toEqual([2, 3]);
expectArraysClose(await da.data(), [0, 0, 0, 0, 0, 0]);
});
it('throws error for string tensor', () => {
expect(() => tf.argMax(['a']))
.toThrowError(/Argument 'x' passed to 'argMax' must be numeric tensor/);
});
});
//# sourceMappingURL=arg_max_test.js.map