Make slice2d internal use makeOutputArray. (tensorflow#124)

* slice 2d => makeOutputArray

scripts/watch-demo

@@ -30,7 +30,7 @@ let httpServerStarted = false;
 
 console.log('Waiting for initial compile...');
 watchify.stderr.on('data', (data) => {
-  if (data.toString().includes(`written to ${path.dirname(startTsFilePath)}`)) {
+  if (data.toString().includes(`bytes written to`)) {
     if (!httpServerStarted) {
       const httpCmd = path.join('node_modules', '.bin', 'http-server');
       const httpServer = spawn(httpCmd, ['-c-1'], { detached: false});

src/math/math_gpu.ts

@@ -75,7 +75,7 @@ export class NDArrayMathGPU extends NDArrayMath {
     // Pretend the source was in logical shape that matches the texture shape.
     const source = ndarray.as2D(texShape[0], texShape[1]);
     // Do the same for output.
-    const output = this.makeOutputArray(texShape) as Array2D;
+    const output = this.makeOutputArray<Array2D>(texShape);
     this.copy2D(source, [0, 0], texShape, output, [0, 0], texShape);
     // Get back to the original logical shape.
     return output.reshape(ndarray.shape);
@@ -84,10 +84,7 @@ export class NDArrayMathGPU extends NDArrayMath {
   protected slice2DInternal(
       input: Array2D, beginRowCol: [number, number],
       sizeRowCol: [number, number]): Array2D {
-    const result = NDArray.make<Array2D>(sizeRowCol, {
-      texture: this.textureManager.acquireTexture(sizeRowCol),
-      textureShapeRC: sizeRowCol
-    });
+    const result = this.makeOutputArray<Array2D>(sizeRowCol);
     this.copy2DInternal(
         input, beginRowCol, sizeRowCol, result, [0, 0], sizeRowCol);
     return result;