pytorch-stuff/Numpy_Bridge.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([1., 1., 1., 1., 1.])\n",
      "[1. 1. 1. 1. 1.]\n",
      "tensor([2., 2., 2., 2., 2.])\n",
      "[2. 2. 2. 2. 2.]\n"
     ]
    }
   ],
   "source": [
    "# the numpy array is just a pointer to the torch tensor\n",
    "a = torch.ones(5)\n",
    "print(a)\n",
    "b = a.numpy()\n",
    "print(b)\n",
    "a.add_(1)\n",
    "print(a)\n",
    "print(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2. 2. 2. 2. 2.]\n",
      "tensor([2., 2., 2., 2., 2.], dtype=torch.float64)\n"
     ]
    }
   ],
   "source": [
    "a = np.ones(5)\n",
    "b = torch.from_numpy(a)\n",
    "np.add(a,1,out=a)\n",
    "print(a)\n",
    "print(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "r2 = torch.randn(4,4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "r = torch.rand(4,4)\n",
    "\n",
    "add_result = torch.add(r,r2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.5124, -0.0635, -1.0141,  0.2371],\n",
      "        [ 0.7809, -1.0184,  1.1734, -0.4778],\n",
      "        [ 1.6340,  0.0201,  0.4821, -0.1986],\n",
      "        [ 1.7512, -0.1389,  1.4546, -0.2230]], device='cuda:0')\n"
     ]
    }
   ],
   "source": [
    "# Move tensor to GPU\n",
    "r2 = r2.cuda()\n",
    "print(r2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n",
      "tensor([[ 0.0573,  0.7549, -0.1884,  0.6424],\n",
      "        [ 0.9001, -0.9460,  2.1332,  0.0502],\n",
      "        [ 1.7473,  0.6463,  0.5804,  0.2859],\n",
      "        [ 2.0273,  0.6695,  2.4453,  0.4559]], device='cuda:0')\n"
     ]
    }
   ],
   "source": [
    "CUDA = torch.cuda.is_available()\n",
    "print(CUDA)\n",
    "\n",
    "if CUDA:\n",
    "    add_result = add_result.cuda()\n",
    "    print(add_result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[-0.4551,  0.6915, -1.2024,  0.8795],\n",
       "        [ 1.6810, -1.9643,  3.3066, -0.4276],\n",
       "        [ 3.3813,  0.6664,  1.0625,  0.0872],\n",
       "        [ 3.7784,  0.5306,  3.8999,  0.2329]], device='cuda:0')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "add_result.add(r2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2, 3, 4, 1]\n",
      "tensor([2, 3, 4, 1]) torch.int64\n"
     ]
    }
   ],
   "source": [
    "a = [2,3,4,1]\n",
    "print(a)\n",
    "to_list = torch.tensor(a)\n",
    "print(to_list, to_list.dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[1, 2],\n",
      "        [3, 4],\n",
      "        [5, 6],\n",
      "        [7, 8]]) torch.int64\n"
     ]
    }
   ],
   "source": [
    "data = [[1,2],[3,4],[5,6],[7,8]]\n",
    "T = torch.tensor(data)\n",
    "print(T,T.dtype)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tensor Concatenation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 0.7607,  0.2490, -1.5199,  0.4037,  0.0063],\n",
      "        [ 0.2223, -1.0452, -1.6327, -0.1692, -0.8291]])\n",
      "tensor([[-0.2285,  0.7535, -1.4712,  1.1518,  0.4560],\n",
      "        [-0.4817, -0.6983, -0.9611, -1.5915, -1.7998],\n",
      "        [-0.3149,  0.4309,  1.4270,  0.1497, -0.4793]])\n",
      "\n",
      "\n",
      "tensor([[ 0.7607,  0.2490, -1.5199,  0.4037,  0.0063],\n",
      "        [ 0.2223, -1.0452, -1.6327, -0.1692, -0.8291],\n",
      "        [-0.2285,  0.7535, -1.4712,  1.1518,  0.4560],\n",
      "        [-0.4817, -0.6983, -0.9611, -1.5915, -1.7998],\n",
      "        [-0.3149,  0.4309,  1.4270,  0.1497, -0.4793]])\n"
     ]
    }
   ],
   "source": [
    "first_1 = torch.randn(2,5)\n",
    "print(first_1)\n",
    "second_1 = torch.randn(3,5)\n",
    "print(second_1)\n",
    "# Concaatenate along the 0 dimension row-wise\n",
    "con_1 = torch.cat([first_1,second_1])\n",
    "print(\"\\n\")\n",
    "print(con_1)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.6601,  1.8097,  0.6295],\n",
      "        [-1.8868, -2.2800,  1.7137]])\n",
      "tensor([[ 1.2576, -0.5680,  1.2772, -0.2566, -2.1952],\n",
      "        [-0.4767, -0.5083, -0.0795, -1.5576,  0.6238]])\n",
      "\n",
      "\n",
      "tensor([[-0.6601,  1.8097,  0.6295,  1.2576, -0.5680,  1.2772, -0.2566, -2.1952],\n",
      "        [-1.8868, -2.2800,  1.7137, -0.4767, -0.5083, -0.0795, -1.5576,  0.6238]])\n"
     ]
    }
   ],
   "source": [
    "first_2 = torch.randn(2,3)\n",
    "print(first_2)\n",
    "second_2 = torch.randn(2,5)\n",
    "print(second_2)\n",
    "# Concaatenate along the 1 dimension column-wise\n",
    "con_2 = torch.cat([first_2,second_2],1)\n",
    "print(\"\\n\")\n",
    "print(con_2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Adding ddimensions to tensors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[2, 3, 4, 5]])\n",
      "torch.Size([1, 4])\n",
      "torch.Size([4])\n"
     ]
    }
   ],
   "source": [
    "tensor_1 = torch.tensor([2,3,4,5])\n",
    "tensor_a = torch.unsqueeze(tensor_1,0)\n",
    "print(tensor_a)\n",
    "print(tensor_a.shape)\n",
    "print(tensor_1.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}