Matlab code: regression with factor variables

<!-- This HTML was auto-generated from MATLAB code. To make modifications, replace the MATLAB code and republish this doc.

Contents

• Draw a scatter plot of boxoffice towards rating, grouped by score
• Create dataset array, convert score to a nominal array
• Queston 1 in Matlab
• Query 2 in Matlab
• Questions three in Matlab
• 20/20

% Apply of Regression with categorical covariates
% By Segovia on 01/03/2014
clear;clc;shut;
load MovieData.mat

Draw a scatter plot of boxoffice towards rating, grouped by score

determine();
gscatter(rating,boxoffice, score,'bgr','x.o^');
title('boxoffice vs. rating, grouped by score')

Create dataset array, convert score to a nominal array

Film=dataset(boxoffice, rating,score);
Film.score=nominal(Film.score);

Queston 1 in Matlab

%Match a regression mannequin
% in 2013a model, use perform LinearModel.match
% in 2013b model, use perform fitlm
match=LinearModel.match(Film, 'rating~score')

match = 


Linear regression mannequin:
    rating ~ 1 + score

Estimated Coefficients:
                    Estimate    SE        tStat      pValue    
    (Intercept)       sixty seven.sixty five     7.1933     9.4046    1.7256e-sixteen
    rating_PG       -12.593     7.8486    -1.6045       zero.11093
    rating_PG-thirteen    -eleven.815     7.4113    -1.5941       zero.11323
    rating_R         -12.02     7.4755    -1.6079       zero.11017


Variety of observations: one hundred forty, Error levels of freedom: 136
Root Imply Squared Error: 14.four
R-squared: zero.0199,  Adjusted R-Squared -zero.00177
F-statistic vs. fixed mannequin: zero.918, p-worth = zero.434

Query 2 in Matlab

%Match a regression mannequin and use "R" as reference degree in score
Movie2=Film;
Movie2.score=reorderlevels(Movie2.score, 'R','G','PG','PG-thirteen');
fit2=LinearModel.match(Movie2,'rating~score')

fit2 = 


Linear regression mannequin:
    rating ~ 1 + score

Estimated Coefficients:
                    Estimate    SE        tStat       pValue    
    (Intercept)        fifty five.sixty three    2.0346      27.342    four.0302e-fifty seven
    rating_G           12.02    7.4755      1.6079       zero.11017
    rating_PG       -zero.57286    three.7411    -zero.15313       zero.87852
    rating_PG-thirteen     zero.20538    2.7062    zero.075893       zero.93962


Variety of observations: one hundred forty, Error levels of freedom: 136
Root Imply Squared Error: 14.four
R-squared: zero.0199,  Adjusted R-Squared -zero.00177
F-statistic vs. fixed mannequin: zero.918, p-worth = zero.434

Questions three in Matlab

anova(match)

ans = 

              SumSq     DF     MeanSq    F          pValue 
    score    570.12      three    one hundred ninety.04    zero.91818    zero.43398
    Error      28149    136    206.ninety eight                      

20/20

[~,~,st]=anova1(Movie2.rating, Movie2.score,'off');
[c,m,h,nms]=multcompare(st,'show','off','ctype','hsd')

c =

    1.0000    2.0000  -31.2248  -12.0200    7.1848
    1.0000    three.0000   -9.0380    zero.5729   10.1837
    1.0000    four.0000   -7.1578   -zero.2054    6.7470
    2.0000    three.0000   -7.5703   12.5929   32.7560
    2.0000    four.0000   -7.2254   eleven.8146   30.8546
    three.0000    four.0000  -10.0553   -zero.7782    eight.4988


m =

   fifty five.6300    2.0346
   sixty seven.6500    7.1933
   fifty five.0571    three.1394
   fifty five.8354    1.7844


h =

     []


nms = 

    'R'
    'G'
    'PG'
    'PG-thirteen'

Revealed with MATLAB® R2013a

-->