{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Raise and check a flag array with numpy\n",
    "> Handle flag array bits with numpy bitwise operations\n",
    "- categories: [jupyter, numpy]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Often to describe data quality of timelines or images, we use array of integers where each of its bit has a specific meaning, so that we can identify what issues affect each data point.\n",
    "\n",
    "For example we have 10 data points, and we assign an array of 8 bits for data quality.\n",
    "Generally `0` means a good data point, any bit raised is sign of some problem in the data, this is more compressed then using different boolean arrays, and allows to make batch `np.bitwise_and` and `np.bitwise_or` operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "flag = np.zeros(10, dtype=np.uint8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The array uses just 8 bits per element"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variable   Type       Data/Info\n",
      "-------------------------------\n",
      "flag       ndarray    10: 10 elems, type `uint8`, 10 bytes\n",
      "np         module     <module 'numpy' from '/ho<...>kages/numpy/__init__.py'>\n"
     ]
    }
   ],
   "source": [
    "%whos"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Raising a bit seems as easy as adding `2**bit` value to the array,\n",
    "for example the 4th bit is 16, so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "flag[2:5] += 2**4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  0, 16, 16, 16,  0,  0,  0,  0,  0], dtype=uint8)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "flag"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The issue is that only works if that bit was `0`, if it was already raised, we would actually zero it and set the higher bit to 1:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "flag[2] += 2**4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  0, 32, 16, 16,  0,  0,  0,  0,  0], dtype=uint8)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "flag"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use bitwise operations\n",
    "\n",
    "Fortunately `numpy` supports bitwise operations that make this easier, see the 2 functions below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def raise_bit_inplace(flag, bit=0):\n",
    "    \"\"\"Raise bit of the flag array in place\n",
    "    \n",
    "    This function modifies the input array,\n",
    "    it also works on slices\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    flag : np.array\n",
    "        flag bit-array, generally unsigned integer\n",
    "    bit : int\n",
    "        bit number to raise\n",
    "    \"\"\"\n",
    "    flag[:] = np.bitwise_or(flag, 2**bit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def raise_bit(flag, bit=0):\n",
    "    \"\"\"Raise bit of the flag array\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    flag : np.array\n",
    "        flag bit-array, generally unsigned integer\n",
    "    bit : int\n",
    "        bit number to raise\n",
    "        \n",
    "    Returns\n",
    "    -------\n",
    "    output_flag : np.array\n",
    "        input array with the requested bit raised\n",
    "    \"\"\"\n",
    "    return np.bitwise_or(flag, 2**bit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def check_bit(flag, bit=0):\n",
    "    \"\"\"Check if bit of the flag array is raised\n",
    "\n",
    "    The output is a boolean array which could\n",
    "    be used for slicing another array.\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    flag : np.array\n",
    "        flag bit-array, generally unsigned integer\n",
    "    bit : int\n",
    "        bit number to check\n",
    "        \n",
    "    Returns\n",
    "    -------\n",
    "    is_raised : bool np.array\n",
    "        True if the bit is raised, False otherwise    \n",
    "    \"\"\"\n",
    "    return np.bitwise_and(flag, int(2**bit)) > 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "is_bit4_raised = check_bit(flag, bit=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False, False, False,  True,  True, False, False, False, False,\n",
       "       False])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "is_bit4_raised"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert np.all(is_bit4_raised[3:5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "They also work with slices of an array:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "raise_bit_inplace(flag[6:], bit=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  0, 32, 16, 16,  0,  2,  2,  2,  2], dtype=uint8)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "flag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Running it twice doesn't change the value of the flag\n",
    "raise_bit_inplace(flag[6:], bit=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  0, 32, 16, 16,  0,  2,  2,  2,  2], dtype=uint8)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "flag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False, False, False, False, False, False,  True,  True,  True,\n",
       "        True])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "check_bit(flag, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First blog post using a Jupyter Notebook with [fastpages](https://github.com/fastai/fastpages#writing-blog-posts-with-jupyter)!!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "PySM 3",
   "language": "python",
   "name": "pysm3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}