concatenatingColumns_nparray.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are a number of methods that allow us to build a 2D `numpy.ndarray` from a number of columns. It is a little harder to append a column to a `numpy.ndarray`. In both cases, however, `np.c_` works. Illustrated below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from numpy.testing.utils import assert_allclose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x1 = np.random.random(size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x2 = np.random.random(size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.98365875  0.72902872  0.91876326  0.69135701  0.39505799  0.16029197\n",
      "  0.91822414  0.35778621  0.96983128  0.60210966]\n"
     ]
    }
   ],
   "source": [
    "print(x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.4477385   0.26924824  0.37417844  0.61372395  0.53000419  0.78225189\n",
      "  0.7915513   0.86612061  0.60639261  0.23050247]\n"
     ]
    }
   ],
   "source": [
    "print(x2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x3 = np.random.random(size=len(x1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.60961559  0.88330827  0.38948932  0.92968805  0.65306924  0.45488228\n",
      "  0.30477829  0.92612084  0.85976936  0.67704528]\n"
     ]
    }
   ],
   "source": [
    "print(x3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To build"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "arr = np.empty(shape=(len(x1),2))\n",
    "arr[:, 0] = x1\n",
    "arr[:, 1] = x2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.98365875  0.4477385 ]\n",
      " [ 0.72902872  0.26924824]\n",
      " [ 0.91876326  0.37417844]\n",
      " [ 0.69135701  0.61372395]\n",
      " [ 0.39505799  0.53000419]\n",
      " [ 0.16029197  0.78225189]\n",
      " [ 0.91822414  0.7915513 ]\n",
      " [ 0.35778621  0.86612061]\n",
      " [ 0.96983128  0.60639261]\n",
      " [ 0.60210966  0.23050247]]\n"
     ]
    }
   ],
   "source": [
    "print(arr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Other methods \n",
    "\n",
    "While all the different methods work on this easily, I was not aware of method1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "method1 = np.c_[x1, x2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "method2 = np.array([x1, x2]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "method3 = np.array(zip(x1, x2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "assert_allclose(arr, method1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "assert_allclose(arr, method2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "assert_allclose(arr, method3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now to build the 2D array with 3 columns\n",
    "\n",
    "Again, doing this from the individual columns is easier than appending the column to the 2 column `numpy.ndarray` by all of the other methods. `method1` is quite nifty here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "arr3c = np.empty(shape=(len(x1), 3))\n",
    "arr3c[:, 0] = x1\n",
    "arr3c[:, 1] = x2\n",
    "arr3c[:, 2] = x3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### From individual columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "method1_3 = np.c_[x1, x2, x3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "method2_3 = np.array([x1, x2, x3]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "method3_3 = np.array(zip(x1, x2, x3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### From the array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "x = np.c_[arr, x3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "assert_allclose(x, arr3c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The others don't work, and probably need a layer of list comprehension"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0.9836587455388892, 0.44773849662426657], 0.6096155922181713],\n",
       "       [[0.7290287205744772, 0.26924823511619966], 0.88330826766873316],\n",
       "       [[0.9187632649518817, 0.37417844298667813], 0.38948931921290275],\n",
       "       [[0.691357011711442, 0.6137239503591148], 0.92968804690670237],\n",
       "       [[0.39505798716349694, 0.5300041889574952], 0.65306923537717032],\n",
       "       [[0.16029196645129973, 0.7822518872170735], 0.45488228196852942],\n",
       "       [[0.9182241436937221, 0.7915512983342922], 0.30477828704136989],\n",
       "       [[0.35778621149379897, 0.8661206058820488], 0.92612083900837938],\n",
       "       [[0.9698312829207162, 0.6063926056387653], 0.85976936161382711],\n",
       "       [[0.6021096554913019, 0.23050246590737478], 0.67704527503701506]], dtype=object)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(zip(arr.tolist(), x3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0.9836587455388892, 0.44773849662426657],\n",
       "        [0.7290287205744772, 0.26924823511619966],\n",
       "        [0.9187632649518817, 0.37417844298667813],\n",
       "        [0.691357011711442, 0.6137239503591148],\n",
       "        [0.39505798716349694, 0.5300041889574952],\n",
       "        [0.16029196645129973, 0.7822518872170735],\n",
       "        [0.9182241436937221, 0.7915512983342922],\n",
       "        [0.35778621149379897, 0.8661206058820488],\n",
       "        [0.9698312829207162, 0.6063926056387653],\n",
       "        [0.6021096554913019, 0.23050246590737478]],\n",
       "       [0.6096155922181713, 0.8833082676687332, 0.38948931921290275,\n",
       "        0.9296880469067024, 0.6530692353771703, 0.4548822819685294,\n",
       "        0.3047782870413699, 0.9261208390083794, 0.8597693616138271,\n",
       "        0.6770452750370151]], dtype=object)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array([arr.tolist(), x3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}