UNPKG

@tensorflow/tfjs-core

Version:

Hardware-accelerated JavaScript library for machine intelligence

495 lines 22.4 kB
/** * @license * Copyright 2017 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 * as tf from '../index'; import { ALL_ENVS, describeWithFlags } from '../jasmine_util'; import { expectArraysClose, expectArraysEqual } from '../test_util'; describeWithFlags('concat1d', ALL_ENVS, () => { it('3 + 5', async () => { const a = tf.tensor1d([3]); const b = tf.tensor1d([5]); const result = tf.concat1d([a, b]); const expected = [3, 5]; expectArraysClose(await result.data(), expected); }); it('TensorLike 3 + 5', async () => { const a = [3]; const b = [5]; const result = tf.concat1d([a, b]); const expected = [3, 5]; expectArraysClose(await result.data(), expected); }); it('TensorLike Chained 3 + 5', async () => { const a = tf.tensor1d([3]); const b = [5]; const result = a.concat([b]); const expected = [3, 5]; expectArraysClose(await result.data(), expected); }); it('3 + [5,7]', async () => { const a = tf.tensor1d([3]); const b = tf.tensor1d([5, 7]); const result = tf.concat1d([a, b]); const expected = [3, 5, 7]; expectArraysClose(await result.data(), expected); }); it('[3,5] + 7', async () => { const a = tf.tensor1d([3, 5]); const b = tf.tensor1d([7]); const result = tf.concat1d([a, b]); const expected = [3, 5, 7]; expectArraysClose(await result.data(), expected); }); it('3 + 5 + 7 + 9', async () => { const a = tf.tensor1d([3]); const b = tf.tensor1d([5]); const c = tf.tensor1d([7]); const d = tf.tensor1d([9]); const result = tf.concat1d([a, b, c, d]); expectArraysClose(await result.data(), [3, 5, 7, 9]); }); it('single tensor', async () => { const a = tf.tensor1d([3]); const result = tf.concat1d([a]); expectArraysClose(await result.data(), [3]); }); it('accepts a tensor-like object', async () => { const a = [3]; const b = [5]; const result = tf.concat1d([a, b]); const expected = [3, 5]; expectArraysClose(await result.data(), expected); }); it('concat complex input', async () => { // [1+1j, 2+2j] const c1 = tf.complex([1, 2], [1, 2]); // [3+3j, 4+4j] const c2 = tf.complex([3, 4], [3, 4]); const axis = 0; const result = tf.concat([c1, c2], axis); const expected = [1, 1, 2, 2, 3, 3, 4, 4]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); }); describeWithFlags('concat2d', ALL_ENVS, () => { it('[[3]] + [[5]], axis=0', async () => { const axis = 0; const a = tf.tensor2d([3], [1, 1]); const b = tf.tensor2d([5], [1, 1]); const result = tf.concat2d([a, b], axis); const expected = [3, 5]; expect(result.shape).toEqual([2, 1]); expectArraysClose(await result.data(), expected); }); it('TensorLike [[3]] + [[5]], axis=0', async () => { const axis = 0; const a = [[3]]; const b = [[5]]; const result = tf.concat2d([a, b], axis); const expected = [3, 5]; expect(result.shape).toEqual([2, 1]); expectArraysClose(await result.data(), expected); }); it('TensorLike Chained [[3]] + [[5]], axis=0', async () => { const axis = 0; const a = tf.tensor2d([3], [1, 1]); const b = [[5]]; const result = a.concat([b], axis); const expected = [3, 5]; expect(result.shape).toEqual([2, 1]); expectArraysClose(await result.data(), expected); }); it('[[3]] + [[5]], axis=1', async () => { const axis = 1; const a = tf.tensor2d([3], [1, 1]); const b = tf.tensor2d([5], [1, 1]); const result = tf.concat2d([a, b], axis); const expected = [3, 5]; expect(result.shape).toEqual([1, 2]); expectArraysClose(await result.data(), expected); }); it('[[1, 2], [3, 4]] + [[5, 6]], axis=0', async () => { const axis = 0; const a = tf.tensor2d([[1, 2], [3, 4]], [2, 2]); const b = tf.tensor2d([[5, 6]], [1, 2]); const result = tf.concat2d([a, b], axis); const expected = [1, 2, 3, 4, 5, 6]; expect(result.shape).toEqual([3, 2]); expectArraysClose(await result.data(), expected); }); it('[[1, 2],[3, 4]] + [[5, 6]] + [[7, 8]], axis=0', async () => { const axis = 0; const a = tf.tensor2d([[1, 2], [3, 4]]); const b = tf.tensor2d([[5, 6]]); const c = tf.tensor2d([[7, 8]]); const result = tf.concat2d([a, b, c], axis); const expected = [1, 2, 3, 4, 5, 6, 7, 8]; expect(result.shape).toEqual([4, 2]); expectArraysClose(await result.data(), expected); }); it('[[1, 2], [3, 4]] + [[5, 6]], axis=1 throws error', () => { const axis = 1; const a = tf.tensor2d([[1, 2], [3, 4]], [2, 2]); const b = tf.tensor2d([[5, 6]], [1, 2]); expect(() => tf.concat2d([a, b], axis)).toThrowError(); }); it('[[1, 2], [3, 4]] + [[5, 6], [7, 8]], axis=1', async () => { const axis = 1; const a = tf.tensor2d([[1, 2], [3, 4]], [2, 2]); const b = tf.tensor2d([[5, 6], [7, 8]], [2, 2]); const result = tf.concat2d([a, b], axis); const expected = [1, 2, 5, 6, 3, 4, 7, 8]; expect(result.shape).toEqual([2, 4]); expectArraysClose(await result.data(), expected); }); it('[[1, 2],[3, 4]] + [[5, 6],[7, 8]] + [[9, 10],[11, 12]], axis=1', async () => { const axis = 1; const a = tf.tensor2d([[1, 2], [3, 4]]); const b = tf.tensor2d([[5, 6], [7, 8]]); const c = tf.tensor2d([[9, 10], [11, 12]]); const result = tf.concat2d([a, b, c], axis); const expected = [1, 2, 5, 6, 9, 10, 3, 4, 7, 8, 11, 12]; expect(result.shape).toEqual([2, 6]); expectArraysClose(await result.data(), expected); }); it('accepts a tensor-like object', async () => { const axis = 0; const a = [[3]]; const b = [[5]]; const result = tf.concat2d([a, b], axis); const expected = [3, 5]; expect(result.shape).toEqual([2, 1]); expectArraysClose(await result.data(), expected); }); it('concat zero-sized tensors', async () => { const a = tf.tensor2d([], [0, 5]); const b = tf.tensor2d([], [0, 5]); const c = tf.tensor2d([], [0, 5]); const res = tf.concat([a, b, c], /* axis */ 0); expect(res.shape).toEqual([0, 5]); expectArraysEqual(await res.data(), []); const res2 = tf.concat([a, b, c], /* axis */ 1); expect(res2.shape).toEqual([0, 15]); expectArraysEqual(await res2.data(), []); }); it('concat complex input axis=0', async () => { // [[1+1j, 2+2j], [3+3j, 4+4j]] const c1 = tf.complex([[1, 2], [3, 4]], [[1, 2], [3, 4]]); // [[5+5j, 6+6j], [7+7j, 8+8j]] const c2 = tf.complex([[5, 6], [7, 8]], [[5, 6], [7, 8]]); const axis = 0; const result = tf.concat([c1, c2], axis); const expected = [1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); it('concat complex input axis=1', async () => { // [[1+1j, 2+2j], [3+3j, 4+4j]] const c1 = tf.complex([[1, 2], [3, 4]], [[1, 2], [3, 4]]); // [[5+5j, 6+6j], [7+7j, 8+8j]] const c2 = tf.complex([[5, 6], [7, 8]], [[5, 6], [7, 8]]); const axis = 1; const result = tf.concat([c1, c2], axis); const expected = [1, 1, 2, 2, 5, 5, 6, 6, 3, 3, 4, 4, 7, 7, 8, 8]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); }); describeWithFlags('concat3d', ALL_ENVS, () => { beforeAll(() => { jasmine.DEFAULT_TIMEOUT_INTERVAL = 1000000; }); it('shapes correct concat axis=-1', async () => { const tensor1 = tf.tensor3d([1, 2, 3], [1, 1, 3]); const tensor2 = tf.tensor3d([4, 5, 6], [1, 1, 3]); const values = tf.concat3d([tensor1, tensor2], -1); expect(values.shape).toEqual([1, 1, 6]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('shapes correct concat axis=0', async () => { const tensor1 = tf.tensor3d([1, 2, 3], [1, 1, 3]); const tensor2 = tf.tensor3d([4, 5, 6], [1, 1, 3]); const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([2, 1, 3]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('concat axis=0', async () => { const tensor1 = tf.tensor3d([1, 11, 111, 2, 22, 222], [1, 2, 3]); const tensor2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([3, 2, 3]); expectArraysClose(await values.data(), [ 1, 11, 111, 2, 22, 222, 5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888 ]); }); it('TensorLike concat axis=0', async () => { const tensor1 = [[[1, 11, 111], [2, 22, 222]]]; const tensor2 = [[[5, 55, 555], [6, 66, 666]], [[7, 77, 777], [8, 88, 888]]]; const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([3, 2, 3]); expectArraysClose(await values.data(), [ 1, 11, 111, 2, 22, 222, 5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888 ]); }); it('Accepts string tensor.', async () => { const tensor1 = tf.tensor3d(['one', 'two', 'three'], [1, 1, 3], 'string'); const tensor2 = tf.tensor3d(['four', 'five', 'six'], [1, 1, 3], 'string'); const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([2, 1, 3]); expectArraysClose(await values.data(), ['one', 'two', 'three', 'four', 'five', 'six']); }); it('TensorLike Chained concat axis=0', async () => { const tensor1 = tf.tensor3d([1, 11, 111, 2, 22, 222], [1, 2, 3]); const tensor2 = [[[5, 55, 555], [6, 66, 666]], [[7, 77, 777], [8, 88, 888]]]; const values = tensor1.concat([tensor2], 0); expect(values.shape).toEqual([3, 2, 3]); expectArraysClose(await values.data(), [ 1, 11, 111, 2, 22, 222, 5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888 ]); }); it('shapes correct concat axis=1', async () => { const tensor1 = tf.tensor3d([1, 2, 3], [1, 1, 3]); const tensor2 = tf.tensor3d([4, 5, 6], [1, 1, 3]); const values = tf.concat3d([tensor1, tensor2], 1); expect(values.shape).toEqual([1, 2, 3]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('concat axis=1', async () => { const tensor1 = tf.tensor3d([1, 11, 111, 3, 33, 333], [2, 1, 3]); const tensor2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); const values = tf.concat3d([tensor1, tensor2], 1); expect(values.shape).toEqual([2, 3, 3]); expectArraysClose(await values.data(), [ 1, 11, 111, 5, 55, 555, 6, 66, 666, 3, 33, 333, 7, 77, 777, 8, 88, 888 ]); }); it('shapes correct concat axis=2', async () => { const tensor1 = tf.tensor3d([1, 2, 3], [1, 1, 3]); const tensor2 = tf.tensor3d([4, 5, 6], [1, 1, 3]); const values = tf.concat3d([tensor1, tensor2], 2); expect(values.shape).toEqual([1, 1, 6]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('concat a large number of tensors, axis=0', async () => { const tensors = []; const expected = []; for (let i = 0; i < 100; i++) { tensors.push(tf.tensor([i], [1])); expected.push(i); } const axis = 0; const res = tf.concat(tensors, axis); expect(res.shape).toEqual([100]); expect(res.dtype).toBe('float32'); expectArraysClose(await res.data(), expected); }); it('concat a large number of tensors, axis=1', async () => { const tensors = []; const expected = []; for (let i = 0; i < 100; i++) { tensors.push(tf.tensor([i], [1, 1])); expected.push(i); } const axis = 1; const res = tf.concat(tensors, axis); expect(res.shape).toEqual([1, 100]); expect(res.dtype).toBe('float32'); expectArraysClose(await res.data(), expected); }); it('concat axis=2', async () => { const tensor1 = tf.tensor3d([1, 11, 2, 22, 3, 33, 4, 44], [2, 2, 2]); const tensor2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); const values = tf.concat3d([tensor1, tensor2], 2); expect(values.shape).toEqual([2, 2, 5]); expectArraysClose(await values.data(), [ 1, 11, 5, 55, 555, 2, 22, 6, 66, 666, 3, 33, 7, 77, 777, 4, 44, 8, 88, 888 ]); }); it('concat throws when invalid non-axis shapes, axis=0', () => { const axis = 0; const x1 = tf.tensor3d([1, 11, 111], [1, 1, 3]); const x2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); expect(() => tf.concat3d([x1, x2], axis)).toThrowError(); }); it('concat throws when invalid non-axis shapes, axis=1', () => { const axis = 1; const x1 = tf.tensor3d([1, 11, 111], [1, 1, 3]); const x2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); expect(() => tf.concat3d([x1, x2], axis)).toThrowError(); }); it('concat throws when invalid non-axis shapes, axis=2', () => { const axis = 2; const x1 = tf.tensor3d([1, 11, 2, 22], [1, 2, 2]); const x2 = tf.tensor3d([5, 55, 555, 6, 66, 666, 7, 77, 777, 8, 88, 888], [2, 2, 3]); expect(() => tf.concat3d([x1, x2], axis)).toThrowError(); }); it('gradient concat axis=0', async () => { const x1 = tf.tensor3d([1, 11, 2, 22], [1, 2, 2]); const x2 = tf.tensor3d([5, 55, 6, 66, 7, 77, 8, 88], [2, 2, 2]); const dy = tf.tensor3d([66, 6, 55, 5, 44, 4, 33, 3, 22, 2, 11, 1], [3, 2, 2]); const axis = 0; const grads = tf.grads((x1, x2) => tf.concat3d([x1, x2], axis)); const [dx1, dx2] = grads([x1, x2], dy); expect(dx1.shape).toEqual(x1.shape); expectArraysClose(await dx1.data(), [66, 6, 55, 5]); expect(dx2.shape).toEqual(x2.shape); expectArraysClose(await dx2.data(), [44, 4, 33, 3, 22, 2, 11, 1]); }); it('gradient with clones', async () => { const x1 = tf.tensor3d([1, 11, 2, 22], [1, 2, 2]); const x2 = tf.tensor3d([5, 55, 6, 66, 7, 77, 8, 88], [2, 2, 2]); const dy = tf.tensor3d([66, 6, 55, 5, 44, 4, 33, 3, 22, 2, 11, 1], [3, 2, 2]); const axis = 0; const grads = tf.grads((x1, x2) => tf.concat3d([x1.clone(), x2.clone()], axis).clone()); const [dx1, dx2] = grads([x1, x2], dy); expect(dx1.shape).toEqual(x1.shape); expectArraysClose(await dx1.data(), [66, 6, 55, 5]); expect(dx2.shape).toEqual(x2.shape); expectArraysClose(await dx2.data(), [44, 4, 33, 3, 22, 2, 11, 1]); }); it('gradient concat axis=1', async () => { const x1 = tf.tensor3d([1, 11, 2, 22], [2, 1, 2]); const x2 = tf.tensor3d([3, 33, 4, 44, 5, 55, 6, 66], [2, 2, 2]); const dy = tf.tensor3d([66, 6, 55, 5, 44, 4, 33, 3, 22, 2, 11, 1], [2, 3, 2]); const axis = 1; const grads = tf.grads((x1, x2) => tf.concat3d([x1, x2], axis)); const [dx1, dx2] = grads([x1, x2], dy); expect(dx1.shape).toEqual(x1.shape); expectArraysClose(await dx1.data(), [66, 6, 33, 3]); expect(dx2.shape).toEqual(x2.shape); expectArraysClose(await dx2.data(), [55, 5, 44, 4, 22, 2, 11, 1]); }); it('gradient concat axis=2', async () => { const x1 = tf.tensor3d([1, 2, 3, 4], [2, 2, 1]); const x2 = tf.tensor3d([5, 55, 6, 66, 7, 77, 8, 88], [2, 2, 2]); const dy = tf.tensor3d([4, 40, 400, 3, 30, 300, 2, 20, 200, 1, 10, 100], [2, 2, 3]); const axis = 2; const grads = tf.grads((x1, x2) => tf.concat3d([x1, x2], axis)); const [dx1, dx2] = grads([x1, x2], dy); expect(dx1.shape).toEqual(x1.shape); expectArraysClose(await dx1.data(), [4, 3, 2, 1]); expect(dx2.shape).toEqual(x2.shape); expectArraysClose(await dx2.data(), [40, 400, 30, 300, 20, 200, 10, 100]); }); it('gradient concat axis=-1', async () => { const x1 = tf.tensor3d([1, 2, 3, 4], [2, 2, 1]); const x2 = tf.tensor3d([5, 55, 6, 66, 7, 77, 8, 88], [2, 2, 2]); const dy = tf.tensor3d([4, 40, 400, 3, 30, 300, 2, 20, 200, 1, 10, 100], [2, 2, 3]); const axis = -1; const grads = tf.grads((x1, x2) => tf.concat3d([x1, x2], axis)); const [dx1, dx2] = grads([x1, x2], dy); expect(dx1.shape).toEqual(x1.shape); expectArraysClose(await dx1.data(), [4, 3, 2, 1]); expect(dx2.shape).toEqual(x2.shape); expectArraysClose(await dx2.data(), [40, 400, 30, 300, 20, 200, 10, 100]); }); it('accepts a tensor-like object', async () => { const tensor1 = [[[1, 2, 3]]]; // 1x1x3 const tensor2 = [[[4, 5, 6]]]; // 1x1x3 const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([2, 1, 3]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('concat tensors with 0 in their shape', async () => { const tensor1 = tf.tensor3d([1, 2, 3, 4, 5, 6], [2, 3, 1]); const tensor2 = tf.tensor3d([], [0, 3, 1]); const values = tf.concat3d([tensor1, tensor2], 0); expect(values.shape).toEqual([2, 3, 1]); expectArraysClose(await values.data(), [1, 2, 3, 4, 5, 6]); }); it('concat complex input axis=0', async () => { // [[[1+1j, 2+2j], [3+3j, 4+4j], [5+5j, 6+6j]]] const c1 = tf.complex([[[1, 2], [3, 4], [5, 6]]], [[[1, 2], [3, 4], [5, 6]]]); // [[[7+7j, 8+8j], [9+9j, 10+10j], [11+11j, 12+12j]]] const c2 = tf.complex([[[7, 8], [9, 10], [11, 12]]], [[[7, 8], [9, 10], [11, 12]]]); const axis = 0; const result = tf.concat([c1, c2], axis); const expected = [ 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, 12, 12 ]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); it('concat complex input axis=1', async () => { // [[[1+1j, 2+2j], [3+3j, 4+4j], [5+5j, 6+6j]]] const c1 = tf.complex([[[1, 2], [3, 4], [5, 6]]], [[[1, 2], [3, 4], [5, 6]]]); // [[[7+7j, 8+8j], [9+9j, 10+10j], [11+11j, 12+12j]]] const c2 = tf.complex([[[7, 8], [9, 10], [11, 12]]], [[[7, 8], [9, 10], [11, 12]]]); const axis = 1; const result = tf.concat([c1, c2], axis); const expected = [ 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, 12, 12 ]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); it('concat complex input axis=1', async () => { // [[[1+1j, 2+2j], [3+3j, 4+4j], [5+5j, 6+6j]]] const c1 = tf.complex([[[1, 2], [3, 4], [5, 6]]], [[[1, 2], [3, 4], [5, 6]]]); // [[[7+7j, 8+8j], [9+9j, 10+10j], [11+11j, 12+12j]]] const c2 = tf.complex([[[7, 8], [9, 10], [11, 12]]], [[[7, 8], [9, 10], [11, 12]]]); const axis = 2; const result = tf.concat([c1, c2], axis); const expected = [ 1, 1, 2, 2, 7, 7, 8, 8, 3, 3, 4, 4, 9, 9, 10, 10, 5, 5, 6, 6, 11, 11, 12, 12 ]; expect(result.dtype).toEqual('complex64'); expectArraysClose(await result.data(), expected); }); }); describeWithFlags('concat throws for non-tensors', ALL_ENVS, () => { it('throws when passed a non-tensor', () => { expect(() => tf.concat([{}])) .toThrowError(/Argument 'tensors\[0\]' passed to 'concat' must be a Tensor/); }); it('accepts a tensor-like object', async () => { const tensor1 = [[[1, 2, 3, 4]]]; // 1x1x4 const tensor2 = [[[4, 5, 6, 7]]]; // 1x1x4 const values = tf.concat([tensor1, tensor2], 0); expect(values.shape).toEqual([2, 1, 4]); expectArraysClose(await values.data(), [1, 2, 3, 4, 4, 5, 6, 7]); }); }); describeWithFlags('memory test', ALL_ENVS, () => { it('returns a new tensor when op is effectively a no-op.', async () => { const a = tf.tensor1d([]); const b = tf.tensor1d([3]); const result = tf.concat([a, b]); a.dispose(); b.dispose(); expectArraysClose(await result.data(), [3]); }); it('ensure no memory leak', async () => { const numTensorsBefore = tf.memory().numTensors; const numDataIdBefore = tf.engine().backend.numDataIds(); const a = tf.tensor1d([]); const b = tf.tensor1d([3]); const result = tf.concat([a, b]); a.dispose(); b.dispose(); result.dispose(); const numTensorsAfter = tf.memory().numTensors; const numDataIdAfter = tf.engine().backend.numDataIds(); expect(numTensorsAfter).toBe(numTensorsBefore); expect(numDataIdAfter).toBe(numDataIdBefore); }); }); //# sourceMappingURL=concat_test.js.map