{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### pokračování fitování polynomy různého stupně\n",
    "\n",
    "**pravé hodnoty** odpovídají polynomu 4. stupně (dominantně kvadratický)\n",
    "\n",
    "testujeme modely pro polynomy 1. - 10. stupně"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7XCIzoAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f89737af908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "from numpy import *\n",
    "from matplotlib import pyplot as pl\n",
    "\n",
    "x=r_[-3:3:20j]\n",
    "sigy=3.\n",
    "tres=[0.5,0.2,7,-0.5,0] #skutecne parametry\n",
    "ytrue=polyval(tres,x)\n",
    "pl.plot(x,ytrue,'k')\n",
    "y=ytrue+random.normal(0,sigy,size=x.shape)\n",
    "pl.plot(x,y,'*')\n",
    "ords=arange(1,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['33.162    0.223',\n",
       " '-4.349    0.223   11.313',\n",
       " '-4.349   -0.897   11.313    0.188',\n",
       " '-0.655   -0.897    7.555    0.188    0.444',\n",
       " '-0.655   -2.988    7.555    1.188    0.444   -0.092',\n",
       " '-0.785   -2.988    7.838    1.188    0.357   -0.092    0.007',\n",
       " '-0.785   -0.601    7.838   -1.065    0.357    0.426    0.007   -0.033',\n",
       " '-0.377   -0.601    6.273   -1.065    1.271    0.426   -0.161   -0.033    0.009',\n",
       " '-0.377    2.066    6.273   -5.299    1.271    2.206   -0.161   -0.306    0.009    0.014']"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res=[polyfit(x,y,i,cov=True) for i in ords]\n",
    "#[[round(p,3) for p in r[0][::-1]] for r in res]\n",
    "[\"   \".join([\"%6.3f\"%p for p in r[0][::-1]]) for r in res]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "směrodatná odchylka koeficientů:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[' 2.865    1.573',\n",
       " ' 0.616    0.225    0.139',\n",
       " ' 0.616    0.568    0.139    0.088',\n",
       " ' 0.407    0.299    0.261    0.046    0.030',\n",
       " ' 0.423    0.552    0.271    0.223    0.031    0.020',\n",
       " ' 0.482    0.535    0.609    0.216    0.173    0.019    0.013',\n",
       " ' 0.498    0.847    0.629    0.645    0.179    0.141    0.013    0.009',\n",
       " ' 0.562    0.843    1.195    0.642    0.620    0.140    0.110    0.009    0.006',\n",
       " ' 0.572    1.214    1.216    1.512    0.631    0.590    0.112    0.088    0.006    0.004']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "errs=[[round(p,5) for p in sqrt(r[1].diagonal()[::-1])/sigy] for r in res]\n",
    "print('směrodatná odchylka koeficientů:')\n",
    "[\"   \".join([\"%6.3f\"%p for p in e]) for e in errs]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### test sumy čtverců reziduí"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "raw_mimetype": "text/x-python"
   },
   "outputs": [],
   "source": [
    "smanu=[]\n",
    "for i in range(1000):\n",
    "    y=ytrue+random.normal(0,sigy,size=x.shape)\n",
    "    res=[polyfit(x,y,i,cov=False) for i in ords]\n",
    "    smanu.append([((y-polyval(r,x))**2).sum() for r in res] )\n",
    "smanu=array(smanu)\n",
    "#siges=sqrt(array(s0)/(20-1-ords))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[2493.5637695312548,\n",
       " 47.639453125000088,\n",
       " 44.588183593750081,\n",
       " 15.22070312500003,\n",
       " 14.202246093750029,\n",
       " 13.122949218750025,\n",
       " 12.132812500000023,\n",
       " 11.146191406250024,\n",
       " 10.05781250000002]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGfSS1DiDXpIaZ9BLUuMMeklq3P8BEJvOyjwqIyYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f89736ddb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ok=pl.hist(smanu[:,5]/sigy**2,20,alpha=0.6)\n",
    "ok=pl.hist(smanu[:,8]/sigy**2,20,alpha=0.6)\n",
    "[stats.chi2.fit(sm/sigy**2, floc=0, fscale=1)[0] for sm in smanu.T]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 138.58103253,    2.8518133 ,    2.83957728,    1.01562264,\n",
       "          1.01592899,    1.01340671,    1.01367193,    1.01469931,\n",
       "          1.00651353])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# redukovany chi2\n",
    "msig=smanu.mean(0)/sigy**2/(20-1-ords)\n",
    "msig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.0758789062500047"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f897367ca20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# rozdeleni rozdilu rezidualni sumy\n",
    "dif=(smanu[:,5]-smanu[:,8])/sigy**2\n",
    "ok=pl.hist(dif,20)\n",
    "ps0=stats.chi2.fit(dif, floc=0, fscale=1)[0]\n",
    "voo=stats.chi2(ps0).pdf(ok[1])\n",
    "voo*=sum(ok[0])/sum(voo)\n",
    "pl.plot(ok[1],voo)\n",
    "ps0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5% kvantil Stud. rozdělení (s kles. poctem stupnu volnosti): \n",
      " [ 1.73406361  1.73960673  1.74588368  1.75305036  1.76131014  1.7709334\n",
      "  1.78228756  1.79588482  1.81246112]\n"
     ]
    }
   ],
   "source": [
    "from scipy import stats\n",
    "print(\"5% kvantil Stud. rozdělení (s kles. poctem stupnu volnosti): \\n\",stats.t(19-ords).isf(0.05))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### statistické významnosti\n",
    "(t-hodnoty pro nulovou hypotézu) jednotlivých parametrů\n",
    "\n",
    "pro dané měření mohou vycházet významné hodnoty i u vyšších momentů"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['12.135    0.234',\n",
       " ' 5.221    1.546   78.509',\n",
       " ' 5.556    4.106   83.540    5.183',\n",
       " ' 3.564    8.895   29.817   11.227   21.776',\n",
       " ' 3.428    5.193   28.675    2.764   20.942    0.454',\n",
       " ' 2.971    4.975   11.563    2.648    3.789    0.435    0.330',\n",
       " ' 2.996    5.756   11.662    3.997    3.822    3.303    0.332    3.273',\n",
       " ' 7.012    7.236    1.017    5.024    8.871    4.152    7.817    4.112    7.819',\n",
       " ' 6.742    6.115    0.977    3.615    8.529    2.605    7.515    2.090    7.516    1.691']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[\"   \".join([\"%6.3f\"%p for p in abs(res[i][0][::-1]/array(errs[i]))]) for i in range(len(res))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAggAAAFnCAYAAADDtrFxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n6VcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xae6dac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "strue=[((ytrue-polyval(r[0],x))**2).sum() for r in res]\n",
    "#sigcom=sqrt(array(s0)/(20-1-ords)) # redukovany chi2\n",
    "pl.semilogy(ords,siges)\n",
    "pl.semilogy(ords,sqrt(array(strue)/(20-1-ords)))\n",
    "pl.legend(['red. chi2','true chi2'])\n",
    "pl.grid()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- (i redukované) chi2 polynomu soustavně klesá - vyšší polynomy \"kopírují šum\" v modelových datech\n",
    "- *pravé* chi2 je rozdíl polynomu daného stupně od *nezašuměné* křivky (střední hodnota velkého množství opakování) -- obvykle je menší než rezidua zahrnující šum\n",
    "\n",
    "správným kritériem je **validace** - chi2 na testovacím vzorku..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "opakovaná simulace = validace\n",
    "-------------\n",
    "numerický experiment můžeme snadno opakovat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from numpy.random import normal\n",
    "def test_poly(x,ytrue,sigy,ords=r_[1:10],rep=1):\n",
    "    '''podle rep:\n",
    "    1: rozdil skutecne funkce a modelu\n",
    "    2: rozdil nove sady dat a modelu\n",
    "    '''\n",
    "    y=ytrue+normal(0,sigy,size=x.shape)\n",
    "    res=[polyfit(x,y,i,cov=True) for i in ords]\n",
    "    if rep==1:\n",
    "        scom=[((ytrue-polyval(r[0],x))**2).sum() for r in res]\n",
    "        return array(scom)/(len(x)-1-ords)\n",
    "    if rep==2:\n",
    "        ycom=ytrue+normal(0,sigy,size=x.shape)\n",
    "        scom=[((ycom-polyval(r[0],x))**2).sum() for r in res]\n",
    "        return array(scom)/(len(x)-1-ords)\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1238.85841607,    17.42292297,    17.68858138,     1.41782944,\n",
       "           1.80194635,     2.20549544,     2.62870594,     3.26606789,\n",
       "           3.93339854])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# provedeme 100 opakování testu\n",
    "fullmat=r_[[test_poly(x,ytrue,2,rep=1) for k in range(100)]]\n",
    "fullmat.mean(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "fullmat2=r_[[test_poly(x,ytrue,2,rep=2) for k in range(100)]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0xb92ae10>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgAAAAF5CAYAAADpvZJuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb78e080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.errorbar(ords[1:],fullmat.mean(0)[1:],fullmat.std(0)[1:]/10.)\n",
    "pl.errorbar(ords[1:],fullmat2.mean(0)[1:],fullmat2.std(0)[1:]/10.)\n",
    "pl.xlabel(\"stupen polynomu\")\n",
    "pl.legend([\"gen. chyba\",\"validace\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### záznam hodnot nafitovaných parametrů"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "apars=[test_poly(x,ytrue,2,rep=0) for k in range(100)] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.08969589, -0.34124228,  5.28565051,  0.24595817,  1.54955129,\n",
       "       -0.07131472, -0.13749907,  0.06867428])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def rep_pars(iord=4):\n",
    "    # vraci prumerne hodnoty fitovanych parametru, teoret. nejistoty a rozptyl parametru\n",
    "    zmat=r_[[a[iord][0] for a in apars]]\n",
    "    cmat=r_[[a[iord][1].diagonal() for a in apars]]\n",
    "    zrep=array([zmat.mean(0),zmat.std(0),sqrt(cmat.mean(0)),sqrt(cmat).std(0)])\n",
    "    return zrep[:,::-1]\n",
    "zot6=rep_pars(6)\n",
    "zot6[0]/zot6[1] # podil stredni hodnoty a nejistoty"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "kovarianční matice dobře odpovídá rozptylu parametrů"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.98661703,  0.95119805,  0.96080738,  0.87037112,  0.95952248,\n",
       "        0.85984801,  0.95468107,  0.86476922])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zot6[1]/zot6[2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### tabulka \"významnosti\" parametrů"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['73.504  3.083',\n",
       " '-5.852  3.083 73.102',\n",
       " '-5.852 -0.720 73.102  2.004',\n",
       " ' 0.024 -0.720 12.865  2.004  8.473',\n",
       " ' 0.024 -0.416 12.865  0.473  8.473 -0.019',\n",
       " ' 0.090 -0.416  5.286  0.473  1.550 -0.019 -0.137',\n",
       " ' 0.090 -0.341  5.286  0.246  1.550 -0.071 -0.137  0.069']"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zotx=[rep_pars(i) for i in range(7)]\n",
    "[' '.join(['%6.3f'%ww for ww in zz[0]/zz[1]]) for zz in zotx]"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}