{ "cells": [ { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "# Перевірка на нормальність" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Більшість методів статистичного аналізу справедливі лише для спостережень, функція розподілу яких підкоряється нормальному закону. Саме тому визначення закону розподілу спостережуваної величини є необхідною складовою будь-якого статистичного аналізу.\n", "\n", "Для більшості класичних методів статистичного аналізу можна знайти аналог, що можна застосовувати і в тих випадках, коли розподіл закону змінної відрізняється від нормального - так звані непараметричні тести. Тим не менше, у своїй більшості непараметричні тести менш точні, чуттєві та надійні. \n", "\n", "Існує кілька способів перевірки аналізованих даних на нормальність розподілу: \n", "1) графічний аналіз (візуальна перевірка), \n", "2) перевірка гіпотези про нормальність розподілу (формальні тести). \n", "\n" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Ми будемо перевіряти на нормальність вагу мишей групи 1 (\"g1\") з файлу mouse.csv" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": false }, "outputs": [ ], "source": [ "dat <- read.csv('mouse.csv')\n", "m1 <- dat$маса[dat$група==\"g1\"] #вага мишей групи 1\n", "m2 <- dat$маса[dat$група==\"g2\"]" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": false }, "outputs": [ ], "source": [ "options(repr.plot.width=5, repr.plot.height=4)" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "### 1) Візуальна перевірка" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Візуальна перевірка може бути використана для оцінки нормальності, хоча цей підхід, як правило, ненадійний і не гарантує, що розподіл є нормальним. Проте, коли дані представлені візуально, читачі статті можуть самостійно судити про припущення щодо нормальності розподілу. \n", "Розподіл частоти (гістограма), графік стовбур-листя (stem-and-leaf plot), boxplot, графік P-P (ймовірність-ймовірність-графік), а також Q-Q графік (quintle-quintle графік) використовуються для перевірки нормальності візуально. \n", "\n", "Гістограма, дає можливість візуально судити про те, чи є розподіл дзвіноподібним, також вона показує наявність прогалини в даних та наявність викідів." ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "##### Гістограма" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Найпростіший графічний спосіб перевірки характеру розподілу даних - побудова гістограми. Якщо гістограма має дзвіноподібний симетричний вид, можна зробити висновок про те, що аналізована змінна має приблизно нормальний розподіл. Однак при інтерпретації гістограм слід дотримуватися обережності, оскільки їх зовнішній вигляд може сильно залежати як від числа спостережень, так і від кроку, обраного для розбиття даних на класи ( hist(...breaks=...) ). Крім того, досить часто при аналізі нормально розподілених але змішаних сукупностей гістограми набувають асиметричний вид, вводячи дослідника в оману.\n" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "побудуємо гістограму та накладем на неї Гаусову криву з середнім та дисперсію як у нашої вибірки m1" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "collapsed": false, "scrolled": true }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "Plot with title “Histogram of Weights for Mouse Group №1”" ] }, "execution_count": 3, "metadata": { }, "output_type": "execute_result" } ], "source": [ "hist(m1,\n", " main=\"Histogram of Weights for Mouse Group №1\",\n", " xlab=\"Mass, g\",\n", " ylab=\"Mass Frequency\",\n", " breaks=15,\n", " freq = F)\n", "\n", "x = seq( min(m1), max(m1), 0.01 )\n", "y = dnorm(x, mean(m1), sd(m1))\n", "lines( x, y, col = \"red\", lwd=2)" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Розглянемо окремі команди" ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "**hist(...., freq = F)** - параметр freq = FALSE дозволяє побудувати гістограму на основі щільності ймовірності, а не частот значень." ] }, { "cell_type": "markdown", "metadata": { "collapsed": false }, "source": [ "Команда **seq( .... )** створює послідовність чисел:" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "