example of sampling_multiple_ordered_seqs

sampling_multiple_ordered_seqs.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-02T00:53:14.764185Z",
     "start_time": "2018-09-02T00:53:14.758997Z"
    }
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "pm = torch.rand(10000)\n",
    "agent_memory_length=5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-02T00:53:14.954330Z",
     "start_time": "2018-09-02T00:53:14.950327Z"
    }
   },
   "outputs": [],
   "source": [
    "batch_start = torch.randint(0, len(pm), (4,)).long()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-02T00:53:37.720369Z",
     "start_time": "2018-09-02T00:53:37.712621Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "batch_indices (multiple ordered sequences) tensor([[ 1022,  1023,  1024,  1025,  1026],\n",
      "        [ 2701,  2702,  2703,  2704,  2705],\n",
      "        [ 4340,  4341,  4342,  4343,  4344],\n",
      "        [ 4538,  4539,  4540,  4541,  4542]])\n"
     ]
    }
   ],
   "source": [
    "batch_indices = torch.stack([torch.arange(batch_start[i], batch_start[i]+agent_memory_length) for i in range(len(batch_start))]).long()\n",
    "print('batch_indices (multiple ordered sequences)', batch_indices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-02T00:53:15.241645Z",
     "start_time": "2018-09-02T00:53:15.234526Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 0.5352,  0.5940,  0.8537,  0.8471,  0.8296],\n",
       "        [ 0.5125,  0.4726,  0.7054,  0.2765,  0.6027],\n",
       "        [ 0.7839,  0.8352,  0.3167,  0.1516,  0.1262],\n",
       "        [ 0.8630,  0.1976,  0.1147,  0.0442,  0.8489]])"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pm[batch_indices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "jupyter3.6",
   "language": "python",
   "name": "jupyter3.6"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "12px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": false,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": true,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}