Sparse Tensor.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sparse Tensor"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Definition"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* A dataset in which most of the entries are zero.\n",
    "* indices: which specifies the indices of the elements in the sparse tensor that contain nonzero values (elements are zero-indexed).\n",
    "* values: which supplies the values for each element in indices.\n",
    "* dense_shape: which specifies the dense_shape of the sparse tensor.\n",
    "* [Document](https://www.tensorflow.org/api_docs/python/tf/sparse/SparseTensor?hl=en)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SparseTensor(indices=tf.Tensor(\n",
      "[[1 1]\n",
      " [3 4]], shape=(2, 2), dtype=int64), values=tf.Tensor([11 56], shape=(2,), dtype=int32), dense_shape=tf.Tensor([5 6], shape=(2,), dtype=int64))\n"
     ]
    }
   ],
   "source": [
    "tensor_sparse = tf.sparse.SparseTensor(\n",
    "    indices=[[1, 1], [3, 4]], values=[11, 56], dense_shape=[5, 6]\n",
    ")\n",
    "\n",
    "print(tensor_sparse)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To Dense Method"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Converts a SparseTensor into a dense tensor.\n",
    "* [Document](https://www.tensorflow.org/api_docs/python/tf/sparse/to_dense?hl=en)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: shape=(5, 6), dtype=int32, numpy=\n",
       "array([[ 0,  0,  0,  0,  0,  0],\n",
       "       [ 0, 11,  0,  0,  0,  0],\n",
       "       [ 0,  0,  0,  0,  0,  0],\n",
       "       [ 0,  0,  0,  0, 56,  0],\n",
       "       [ 0,  0,  0,  0,  0,  0]])>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.sparse.to_dense(tensor_sparse)"
   ]
  }
 ],
 "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": "3.10.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}