@tensorflow/tfjs-core
Version:
Hardware-accelerated JavaScript library for machine intelligence
1,097 lines (1,096 loc) • 155 kB
JavaScript
"use strict";
/**
* @license
* Copyright 2017 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.
* =============================================================================
*/
var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
return new (P || (P = Promise))(function (resolve, reject) {
function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
function step(result) { result.done ? resolve(result.value) : new P(function (resolve) { resolve(result.value); }).then(fulfilled, rejected); }
step((generator = generator.apply(thisArg, _arguments || [])).next());
});
};
var __generator = (this && this.__generator) || function (thisArg, body) {
var _ = { label: 0, sent: function() { if (t[0] & 1) throw t[1]; return t[1]; }, trys: [], ops: [] }, f, y, t, g;
return g = { next: verb(0), "throw": verb(1), "return": verb(2) }, typeof Symbol === "function" && (g[Symbol.iterator] = function() { return this; }), g;
function verb(n) { return function (v) { return step([n, v]); }; }
function step(op) {
if (f) throw new TypeError("Generator is already executing.");
while (_) try {
if (f = 1, y && (t = op[0] & 2 ? y["return"] : op[0] ? y["throw"] || ((t = y["return"]) && t.call(y), 0) : y.next) && !(t = t.call(y, op[1])).done) return t;
if (y = 0, t) op = [op[0] & 2, t.value];
switch (op[0]) {
case 0: case 1: t = op; break;
case 4: _.label++; return { value: op[1], done: false };
case 5: _.label++; y = op[1]; op = [0]; continue;
case 7: op = _.ops.pop(); _.trys.pop(); continue;
default:
if (!(t = _.trys, t = t.length > 0 && t[t.length - 1]) && (op[0] === 6 || op[0] === 2)) { _ = 0; continue; }
if (op[0] === 3 && (!t || (op[1] > t[0] && op[1] < t[3]))) { _.label = op[1]; break; }
if (op[0] === 6 && _.label < t[1]) { _.label = t[1]; t = op; break; }
if (t && _.label < t[2]) { _.label = t[2]; _.ops.push(op); break; }
if (t[2]) _.ops.pop();
_.trys.pop(); continue;
}
op = body.call(thisArg, _);
} catch (e) { op = [6, e]; y = 0; } finally { f = t = 0; }
if (op[0] & 5) throw op[1]; return { value: op[0] ? op[1] : void 0, done: true };
}
};
var _this = this;
Object.defineProperty(exports, "__esModule", { value: true });
var tf = require("../index");
var jasmine_util_1 = require("../jasmine_util");
var test_util_1 = require("../test_util");
var reduce_util = require("./reduce_util");
jasmine_util_1.describeWithFlags('min', jasmine_util_1.ALL_ENVS, function () {
it('Tensor1D', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([3, -1, 0, 100, -7, 2]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.min(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), -7]);
return [2 /*return*/];
}
});
}); });
it('ignores NaNs', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([3, NaN, 2]);
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, tf.min(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), 2]);
return [2 /*return*/];
}
});
}); });
it('2D', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.min(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), -7]);
return [2 /*return*/];
}
});
}); });
it('2D axis=[0,1]', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.min(a, [0, 1]).data()];
case 1:
_a.apply(void 0, [_b.sent(), -7]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=0', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
r = tf.min(a, 0);
expect(r.shape).toEqual([3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [3, -7, 0]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=0, keepDims', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
r = tf.min(a, 0, true /* keepDims */);
expect(r.shape).toEqual([1, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [3, -7, 0]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=1 provided as a number', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.min(a, 1);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [2, -7]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis = -1 provided as a number', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.min(a, -1);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [2, -7]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=[1]', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.min(a, [1]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [2, -7]]);
return [2 /*return*/];
}
});
}); });
it('throws when passed a non-tensor', function () {
expect(function () { return tf.min({}); })
.toThrowError(/Argument 'x' passed to 'min' must be a Tensor/);
});
it('accepts a tensor-like object', function () { return __awaiter(_this, void 0, void 0, function () {
var _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.min([3, -1, 0, 100, -7, 2]).data()];
case 1:
_a.apply(void 0, [_b.sent(), -7]);
return [2 /*return*/];
}
});
}); });
it('min gradient: Scalar', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.scalar(42);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.min(v); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), -1]);
return [2 /*return*/];
}
});
}); });
it('gradient with clones', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.scalar(42);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.min(v.clone()).clone(); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), -1]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 1D, ties', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor1d([-1, -3, -7, -7]);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.min(v); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, 0, -1, -1]]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 2D, axes=-1, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[-0, -20, -10], [10, 30, 20]]);
dy = tf.tensor1d([-1, -1]);
axis = -1;
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, 0, 0]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('min gradient: ties, 2D, axes=-1, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, -20, -20], [10, 30, 10]]);
dy = tf.tensor1d([-1, -1]);
axis = -1;
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, -1, -1, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 2D, axes=0, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 10], [-10, -30, 20]]);
dy = tf.tensor1d([-1, -1, -1]);
axis = 0;
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1, -1, 0, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 2D, axes=-1, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, -20, -10], [10, 30, 20]]);
dy = tf.tensor2d([[-1], [-1]]);
axis = -1;
keepDims = true;
gradients = tf.grad(function (v) { return tf.min(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, 0, 0]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 2D, axes=0, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, -20, -10], [10, 30, -20]]);
dy = tf.tensor2d([[-1, -1, -1]]);
axis = 0;
keepDims = true;
gradients = tf.grad(function (v) { return tf.min(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1, -1, 0, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 3D, axes=[1, 2], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, -20], [-10, -15]], [[10, 30], [20, 15]]]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2];
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, 0, -1, 0, 0, 0]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('min gradient: ties, 3D, axes=[1, 2], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, -20], [-20, -20]], [[10, 30], [10, 15]]]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2];
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, -1, -1, -1, 0, -1, 0]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 3D, axes=2, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, -20], [-10, -15]], [[10, 30], [20, 15]]]);
dy = tf.tensor2d([[-1, -1], [-1, -1]]);
axis = 2;
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, -1, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('min gradient: 3D, axes=2, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, -20], [-10, -15]], [[10, 30], [20, 15]]]);
dy = tf.tensor3d([[[-1], [-1]], [[-1], [-1]]]);
axis = 2;
keepDims = true;
gradients = tf.grad(function (v) { return tf.min(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, -1, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('min gradient: ties, 4D, axes=[1, 2, 3], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor4d([
[[[0, -20], [-20, -20]], [[10, 30], [10, 30]]],
[[[0, 20], [20, 20]], [[-10, -30], [-10, -30]]]
]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2, 3];
gradients = tf.grad(function (v) { return tf.min(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(),
[0, -1, -1, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('min gradient: ties, 4D, axes=[2, 3], keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor4d([
[[[0, -20], [-20, -20]], [[10, 30], [10, 30]]],
[[[0, 20], [20, 20]], [[-10, -30], [-10, -30]]]
]);
dy = tf.tensor4d([[[[-1]], [[-2]]], [[[-3]], [[-4]]]]);
axis = [2, 3];
keepDims = true;
gradients = tf.grad(function (v) { return tf.min(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(),
[0, -1, -1, -1, -2, 0, -2, 0, -3, 0, 0, 0, 0, -4, 0, -4]]);
expect(gradients.shape).toEqual([2, 2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('throws error for string tensor', function () {
expect(function () { return tf.min(['a']); })
.toThrowError(/Argument 'x' passed to 'min' must be numeric tensor/);
});
});
jasmine_util_1.describeWithFlags('max', jasmine_util_1.ALL_ENVS, function () {
it('with one element dominating', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([3, -1, 0, 100, -7, 2]);
r = tf.max(a);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), 100]);
return [2 /*return*/];
}
});
}); });
it('with all elements being the same', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([3, 3, 3]);
r = tf.max(a);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), 3]);
return [2 /*return*/];
}
});
}); });
it('ignores NaNs', function () { return __awaiter(_this, void 0, void 0, function () {
var _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.max([3, NaN, 2]).data()];
case 1:
_a.apply(void 0, [_b.sent(), 3]);
return [2 /*return*/];
}
});
}); });
it('2D', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.max(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), 100]);
return [2 /*return*/];
}
});
}); });
it('2D axis=[0,1]', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, tf.max(a, [0, 1]).data()];
case 1:
_a.apply(void 0, [_b.sent(), 100]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=0', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
r = tf.max(a, [0]);
expect(r.shape).toEqual([3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [100, -1, 2]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=0, keepDims', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
r = tf.max(a, [0], true /* keepDims */);
expect(r.shape).toEqual([1, 3]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [100, -1, 2]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=1 provided as a number', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.max(a, 1);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [5, 100]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis = -1 provided as a number', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.max(a, -1);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [5, 100]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=[1]', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, 2, 5, 100, -7, 2], [2, 3]);
r = tf.max(a, [1]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [5, 100]]);
return [2 /*return*/];
}
});
}); });
it('6D, axis=[5]', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, expectedResult, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.range(0, 64).reshape([2, 2, 2, 2, 2, 2]);
r = tf.max(a, [5]);
expectedResult = [
1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31,
33, 35, 37, 39, 41, 43, 45, 47, 49, 51, 53, 55, 57, 59, 61, 63
];
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), expectedResult]);
return [2 /*return*/];
}
});
}); });
it('throws when passed a non-tensor', function () {
expect(function () { return tf.max({}); })
.toThrowError(/Argument 'x' passed to 'max' must be a Tensor/);
});
it('accepts a tensor-like object', function () { return __awaiter(_this, void 0, void 0, function () {
var r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
r = tf.max([3, -1, 0, 100, -7, 2]);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), 100]);
return [2 /*return*/];
}
});
}); });
it('max gradient: Scalar', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.scalar(42);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.max(v); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1]]);
return [2 /*return*/];
}
});
}); });
it('gradient with clones', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.scalar(42);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.max(v.clone()).clone(); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1]]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 1D, ties', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor1d([1, 3, 7, 7]);
dy = tf.scalar(-1);
gradients = tf.grad(function (v) { return tf.max(v); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, 0, -1, -1]]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 2D, axes=-1, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 10], [-10, -30, -20]]);
dy = tf.tensor1d([-1, -1]);
axis = -1;
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, 0, 0]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('max gradient: ties, 2D, axes=-1, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 20], [-10, -30, -10]]);
dy = tf.tensor1d([-1, -1]);
axis = -1;
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, -1, -1, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 2D, axes=0, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 10], [-10, -30, 20]]);
dy = tf.tensor1d([-1, -1, -1]);
axis = 0;
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1, -1, 0, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 2D, axes=-1, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 10], [-10, -30, -20]]);
dy = tf.tensor2d([[-1], [-1]]);
axis = -1;
keepDims = true;
gradients = tf.grad(function (v) { return tf.max(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, 0, 0]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 2D, axes=0, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor2d([[0, 20, 10], [-10, -30, 20]]);
dy = tf.tensor2d([[-1, -1, -1]]);
axis = 0;
keepDims = true;
gradients = tf.grad(function (v) { return tf.max(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [-1, -1, 0, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 3]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 3D, axes=[1, 2], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, 20], [10, 15]], [[-10, -30], [-20, -15]]]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2];
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, 0, -1, 0, 0, 0]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('max gradient: ties, 3D, axes=[1, 2], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, 20], [20, 20]], [[-10, -30], [-10, -15]]]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2];
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, -1, -1, -1, 0, -1, 0]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 3D, axes=2, keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, 20], [10, 15]], [[-10, -30], [-20, -15]]]);
dy = tf.tensor2d([[-1, -1], [-1, -1]]);
axis = 2;
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, -1, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('max gradient: 3D, axes=2, keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor3d([[[0, 20], [10, 15]], [[-10, -30], [-20, -15]]]);
dy = tf.tensor3d([[[-1], [-1]], [[-1], [-1]]]);
axis = 2;
keepDims = true;
gradients = tf.grad(function (v) { return tf.max(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(), [0, -1, 0, -1, -1, 0, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('max gradient: ties, 4D, axes=[1, 2, 3], keepDims=false', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor4d([
[[[0, 20], [20, 20]], [[-10, -30], [-10, -30]]],
[[[0, -20], [-20, -20]], [[10, 30], [10, 30]]]
]);
dy = tf.tensor1d([-1, -1]);
axis = [1, 2, 3];
gradients = tf.grad(function (v) { return tf.max(v, axis); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(),
[0, -1, -1, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 0, -1]]);
expect(gradients.shape).toEqual([2, 2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('max gradient: ties, 4D, axes=[2, 3], keepDims=true', function () { return __awaiter(_this, void 0, void 0, function () {
var x, dy, axis, keepDims, gradients, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
x = tf.tensor4d([
[[[0, 20], [20, 20]], [[-10, -30], [-10, -30]]],
[[[0, -20], [-20, -20]], [[10, 30], [10, 30]]]
]);
dy = tf.tensor4d([[[[-1]], [[-2]]], [[[-3]], [[-4]]]]);
axis = [2, 3];
keepDims = true;
gradients = tf.grad(function (v) { return tf.max(v, axis, keepDims); })(x, dy);
_a = test_util_1.expectArraysClose;
return [4 /*yield*/, gradients.data()];
case 1:
_a.apply(void 0, [_b.sent(),
[0, -1, -1, -1, -2, 0, -2, 0, -3, 0, 0, 0, 0, -4, 0, -4]]);
expect(gradients.shape).toEqual([2, 2, 2, 2]);
return [2 /*return*/];
}
});
}); });
it('throws error for string tensor', function () {
expect(function () { return tf.max(['a']); })
.toThrowError(/Argument 'x' passed to 'max' must be numeric tensor/);
});
});
jasmine_util_1.describeWithFlags('argmax', jasmine_util_1.ALL_ENVS, function () {
it('Tensor1D', function () { return __awaiter(_this, void 0, void 0, function () {
var a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([1, 0, 3, 2]);
result = tf.argMax(a);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), 2]);
return [2 /*return*/];
}
});
}); });
it('one value', function () { return __awaiter(_this, void 0, void 0, function () {
var a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([10]);
result = tf.argMax(a);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), 0]);
return [2 /*return*/];
}
});
}); });
it('N > than parallelization threshold', function () { return __awaiter(_this, void 0, void 0, function () {
var n, values, i, a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
n = reduce_util.PARALLELIZE_THRESHOLD * 2;
values = new Float32Array(n);
for (i = 0; i < n; i++) {
values[i] = i;
}
a = tf.tensor1d(values);
result = tf.argMax(a);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), n - 1]);
return [2 /*return*/];
}
});
}); });
it('3D, N > than parallelization threshold', function () { return __awaiter(_this, void 0, void 0, function () {
var n, values, i, a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
n = reduce_util.PARALLELIZE_THRESHOLD * 2;
values = new Float32Array(n);
for (i = 0; i < n; i++) {
values[i] = i;
}
a = tf.tensor3d(values, [1, 1, n]);
result = tf.argMax(a, -1);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), n - 1]);
return [2 /*return*/];
}
});
}); });
it('max index corresponds to start of a non-initial window', function () { return __awaiter(_this, void 0, void 0, function () {
var n, windowSize, values, index, a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
n = reduce_util.PARALLELIZE_THRESHOLD * 2;
windowSize = reduce_util.computeOptimalWindowSize(n);
values = new Float32Array(n);
index = windowSize * 2;
values[index] = 1;
a = tf.tensor1d(values);
result = tf.argMax(a);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), index]);
return [2 /*return*/];
}
});
}); });
it('5D, max index corresponds to start of a non-initial window', function () { return __awaiter(_this, void 0, void 0, function () {
var n, windowSize, values, index, a, result, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
n = reduce_util.PARALLELIZE_THRESHOLD * 2;
windowSize = reduce_util.computeOptimalWindowSize(n);
values = new Float32Array(n);
index = windowSize * 2;
values[index] = 1;
a = tf.tensor5d(values, [1, 1, 1, 1, n]);
result = tf.argMax(a, -1);
expect(result.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, result.data()];
case 1:
_a.apply(void 0, [_b.sent(), index]);
return [2 /*return*/];
}
});
}); });
it('ignores NaNs', function () { return __awaiter(_this, void 0, void 0, function () {
var a, res, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor1d([0, 3, 5, NaN, 3]);
res = tf.argMax(a);
expect(res.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, res.data()];
case 1:
_a.apply(void 0, [_b.sent(), 2]);
return [2 /*return*/];
}
});
}); });
it('2D, no axis specified', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, tf.argMax(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), [1, 0, 1]]);
return [2 /*return*/];
}
});
}); });
it('4D, no axis specified', function () { return __awaiter(_this, void 0, void 0, function () {
var a, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor4d([3, -1, 0, 100, -7, 2], [2, 1, 1, 3]);
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, tf.argMax(a).data()];
case 1:
_a.apply(void 0, [_b.sent(), [1, 0, 1]]);
return [2 /*return*/];
}
});
}); });
it('2D, axis=0', function () { return __awaiter(_this, void 0, void 0, function () {
var a, r, _a;
return __generator(this, function (_b) {
switch (_b.label) {
case 0:
a = tf.tensor2d([3, -1, 0, 100, -7, 2], [2, 3]);
r = tf.argMax(a, 0);
expect(r.shape).toEqual([3]);
expect(r.dtype).toBe('int32');
_a = test_util_1.expectArraysEqual;
return [4 /*yield*/, r.data()];
case 1:
_a.apply(void 0, [_b.sent(), [1, 0, 1]]);
return [2 /*return*/];