{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Torch tensors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([2, 2, 1])\n"
     ]
    }
   ],
   "source": [
    "# This is a 1D tensor\n",
    "a = torch.tensor([2,2,1])\n",
    "print(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 2,  3,  4],\n",
      "        [ 1,  2,  3],\n",
      "        [ 0,  8,  7],\n",
      "        [10, 11, 12]])\n"
     ]
    }
   ],
   "source": [
    "# This is a 2D tensor\n",
    "b = torch.tensor([[2,3,4],[1,2,3],[0,8,7],[10,11,12]])\n",
    "print(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([3])\n",
      "torch.Size([3])\n",
      "torch.Size([4, 3])\n",
      "torch.Size([4, 3])\n"
     ]
    }
   ],
   "source": [
    "# the size of a tensor\n",
    "print(a.shape)\n",
    "print(a.size())\n",
    "print(b.shape)\n",
    "print(b.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    }
   ],
   "source": [
    "# get the number of rows\n",
    "print(b.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = torch.FloatTensor([[2,3,4],[1,2,3],[0,8,7],[10,11,12]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = torch.DoubleTensor([[2,3,4],[1,2,3],[0,8,7],[10,11,12]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 2.,  3.,  4.],\n",
      "        [ 1.,  2.,  3.],\n",
      "        [ 0.,  8.,  7.],\n",
      "        [10., 11., 12.]])\n",
      "torch.float32\n"
     ]
    }
   ],
   "source": [
    "print(c)\n",
    "print(c.dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 2.,  3.,  4.],\n",
      "        [ 1.,  2.,  3.],\n",
      "        [ 0.,  8.,  7.],\n",
      "        [10., 11., 12.]], dtype=torch.float64)\n",
      "torch.float64\n"
     ]
    }
   ],
   "source": [
    "print(d)\n",
    "print(d.dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(5.2500)\n"
     ]
    }
   ],
   "source": [
    "print(c.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(5.2500, dtype=torch.float64)\n"
     ]
    }
   ],
   "source": [
    "print(d.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(4.1588)\n"
     ]
    }
   ],
   "source": [
    "print(c.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(4.1588, dtype=torch.float64)\n"
     ]
    }
   ],
   "source": [
    "print(d.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 2],\n",
      "        [ 3],\n",
      "        [ 4],\n",
      "        [ 1],\n",
      "        [ 2],\n",
      "        [ 3],\n",
      "        [ 0],\n",
      "        [ 8],\n",
      "        [ 7],\n",
      "        [10],\n",
      "        [11],\n",
      "        [12]])\n",
      "tensor([ 2,  3,  4,  1,  2,  3,  0,  8,  7, 10, 11, 12])\n",
      "tensor([[ 2,  3,  4,  1],\n",
      "        [ 2,  3,  0,  8],\n",
      "        [ 7, 10, 11, 12]])\n",
      "tensor([[ 2,  3,  4,  1,  2,  3,  0,  8,  7, 10, 11, 12]])\n",
      "torch.Size([1, 12])\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "tensor([[[ 2.3594,  0.3172, -2.0944,  0.5019],\n",
      "         [ 0.4300,  0.4082, -0.1532,  0.6677],\n",
      "         [-1.0891,  1.3434, -0.2586, -1.2845]],\n",
      "\n",
      "        [[-2.2463, -0.6665,  0.5781,  1.1152],\n",
      "         [-1.9292, -0.5129, -0.2301, -0.1059],\n",
      "         [-0.5839,  2.9470, -0.6610,  1.1021]]])\n",
      "tensor([[ 2.3594,  0.3172, -2.0944,  0.5019,  0.4300,  0.4082, -0.1532,  0.6677,\n",
      "         -1.0891,  1.3434, -0.2586, -1.2845],\n",
      "        [-2.2463, -0.6665,  0.5781,  1.1152, -1.9292, -0.5129, -0.2301, -0.1059,\n",
      "         -0.5839,  2.9470, -0.6610,  1.1021]])\n",
      "tensor([[ 2.3594,  0.3172, -2.0944,  0.5019,  0.4300,  0.4082, -0.1532,  0.6677,\n",
      "         -1.0891,  1.3434, -0.2586, -1.2845],\n",
      "        [-2.2463, -0.6665,  0.5781,  1.1152, -1.9292, -0.5129, -0.2301, -0.1059,\n",
      "         -0.5839,  2.9470, -0.6610,  1.1021]])\n"
     ]
    }
   ],
   "source": [
    "print(b.view(-1,1))\n",
    "print(b.view(12))\n",
    "print(b.view(-1,4))\n",
    "\n",
    "b = b.view(1,-1)\n",
    "print(b)\n",
    "print(b.shape)\n",
    "print(\"\\n\")\n",
    "#this is the format of a Tensor (channels,rows,columns)\n",
    "three_dim = torch.randn(2,3,4)\n",
    "print(\"\\n\")\n",
    "print(three_dim)\n",
    "print(three_dim.view(2,12))\n",
    "print(three_dim.view(2,-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[0.8131, 0.3529, 0.4527, 0.8643],\n",
      "        [0.5623, 0.7266, 0.7129, 0.4911],\n",
      "        [0.3165, 0.8044, 0.8082, 0.4713],\n",
      "        [0.7482, 0.7304, 0.9080, 0.0568]])\n"
     ]
    }
   ],
   "source": [
    "r = torch.rand(4,4)\n",
    "print(r)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 0.0370, -0.3710,  0.8986, -1.1703],\n",
      "        [-0.8784, -0.7126,  0.4817, -0.9874],\n",
      "        [ 0.2186, -1.7692,  1.3162, -1.2127],\n",
      "        [ 0.2576, -0.4097, -1.5554,  0.2979]])\n",
      "torch.float32\n"
     ]
    }
   ],
   "source": [
    "r2 = torch.randn(4,4)\n",
    "print(r2)\n",
    "print(r2.dtype)"
   ]
  },
  {
   "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
}