{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bb4ff0cf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[0., 0., 0.],\n",
      "        [0., 0., 0.]])\n",
      "tensor([[1., 1., 1.],\n",
      "        [1., 1., 1.]])\n",
      "tensor([[-0.6469,  0.8159, -0.0863],\n",
      "        [ 0.4979, -0.2686, -0.6490]])\n",
      "tensor([[1, 2],\n",
      "        [3, 4]])\n",
      "tensor([[-0.5335,  0.5290, -0.9839],\n",
      "        [ 0.7704,  0.4273,  1.2004]])\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "# 创建一个 2*3 的全 0 张量\n",
    "a = torch.zeros(2, 3)\n",
    "print(a)\n",
    "\n",
    "# 创建一个 2*3 的全 1 张量\n",
    "b = torch.ones(2, 3)\n",
    "print(b)\n",
    "\n",
    "# 创建一个 2*3 的随机张量\n",
    "c = torch.randn(2, 3)\n",
    "print(c)\n",
    "\n",
    "# 从 NumPy 数组创建张量\n",
    "import numpy as np\n",
    "numpy_array = np.array([[1,2],[3,4]])\n",
    "tensor_from_numpy = torch.from_numpy(numpy_array)\n",
    "print(tensor_from_numpy)\n",
    "\n",
    "device  = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "d = torch.randn(2,3, device=device)\n",
    "print(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4bca6c60",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 0.6426,  1.6234, -0.6121],\n",
      "        [ 0.6447, -0.8871, -0.9659]])\n",
      "tensor([[ 0.8695, -0.4598, -0.3521],\n",
      "        [-0.3847,  0.4315,  0.2879]])\n",
      "tensor([[ 1.5121,  1.1636, -0.9642],\n",
      "        [ 0.2600, -0.4556, -0.6781]])\n"
     ]
    }
   ],
   "source": [
    "# 张量相加\n",
    "e = torch.randn(2, 3)\n",
    "f = torch.randn(2, 3)\n",
    "print(e)\n",
    "print(f)\n",
    "print(e + f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b1bfa2d7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 0.5588, -0.7465,  0.2155],\n",
      "        [-0.2480, -0.3828, -0.2780]])\n"
     ]
    }
   ],
   "source": [
    "# 逐元素乘法\n",
    "print(e * f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1e201604",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.6453, -0.7084, -0.4454],\n",
      "        [-0.8929,  0.6103,  1.3243]])\n"
     ]
    }
   ],
   "source": [
    "# 张量的转置\n",
    "g = torch.randn(3, 2)\n",
    "print(g.t())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "2d42353d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([3, 2])\n"
     ]
    }
   ],
   "source": [
    "# 张量的形状\n",
    "print(g.shape)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "shuhe_xwdev (3.11.14)",
   "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": "3.11.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}