{"id":2946,"date":"2023-07-19T23:21:19","date_gmt":"2023-07-19T23:21:19","guid":{"rendered":"https:\/\/statorials.org\/pt\/prever-regressao-multipla-em-r\/"},"modified":"2023-07-19T23:21:19","modified_gmt":"2023-07-19T23:21:19","slug":"prever-regressao-multipla-em-r","status":"publish","type":"post","link":"https:\/\/statorials.org\/pt\/prever-regressao-multipla-em-r\/","title":{"rendered":"Como prever valores em r usando um modelo de regress\u00e3o m\u00faltipla"},"content":{"rendered":"

<\/p>\n

\n

Voc\u00ea pode usar a seguinte sintaxe b\u00e1sica para prever valores em R usando um modelo de regress\u00e3o linear m\u00faltipla ajustado:<\/span><\/span><\/p>\n

 #define new observation\n<\/span>new <- data. frame<\/span> (x1=c(5), x2=c(10), x3=c(12.5))\n\n#use fitted model to predict the response value for the new observation\n<\/span>predict(model, newdata=new)<\/strong><\/pre>\n
 O exemplo a seguir mostra como usar esta fun\u00e7\u00e3o na pr\u00e1tica.<\/span><\/p>\n
 Exemplo: previs\u00e3o de valores usando um modelo de regress\u00e3o linear m\u00faltipla ajustado<\/strong><\/span><\/h3>\n
 Suponha que temos o seguinte conjunto de dados em R que cont\u00e9m informa\u00e7\u00f5es sobre jogadores de basquete:<\/span><\/p>\n
 #create data frame\n<\/span>df <- data. frame<\/span> (rating=c(67, 75, 79, 85, 90, 96, 97),\n                 points=c(8, 12, 16, 15, 22, 28, 24),\n                 assists=c(4, 6, 6, 5, 3, 8, 7),\n                 rebounds=c(1, 4, 3, 3, 2, 6, 7))\n\n#view data frame\n<\/span>df\n\n  rating points assists rebounds\n1 67 8 4 1\n2 75 12 6 4\n3 79 16 6 3\n4 85 15 5 3\n5 90 22 3 2\n6 96 28 8 6\n7 97 24 7 7\n<\/strong><\/pre>\n
 Agora, suponha que ajustamos um modelo de regress\u00e3o linear m\u00faltipla usando pontos<\/strong> , assist\u00eancias<\/strong> e rebotes<\/strong> como vari\u00e1veis preditoras e classifica\u00e7\u00e3o<\/strong> como vari\u00e1vel de resposta<\/a> :<\/span><\/p>\n
 #fit multiple linear regression model\n<\/span>model <- lm(rating ~ points + assists + rebounds, data=df)\n\n#view model summary\n<\/span>summary(model)\n\nCall:\nlm(formula = rating ~ points + assists + rebounds, data = df)\n\nResiduals:\n      1 2 3 4 5 6 7 \n-1.5902 -1.7181 0.2413 4.8597 -1.0201 -0.6082 -0.1644 \n\nCoefficients:\n            Estimate Std. Error t value Pr(>|t|)   \n(Intercept) 66.4355 6.6932 9.926 0.00218 **\npoints 1.2152 0.2788 4.359 0.02232 * \nassists -2.5968 1.6263 -1.597 0.20860   \nrebounds 2.8202 1.6118 1.750 0.17847   \n---\nSignificant. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nResidual standard error: 3.193 on 3 degrees of freedom\nMultiple R-squared: 0.9589, Adjusted R-squared: 0.9179 \nF-statistic: 23.35 on 3 and 3 DF, p-value: 0.01396\n<\/strong><\/pre>\n
 A partir dos valores da coluna Estimativa<\/strong> , podemos escrever o modelo de regress\u00e3o ajustado:<\/span><\/p>\n
 Pontua\u00e7\u00e3o = 66,4355 + 1,2151 (pontos) \u2013 2,5968 (assist\u00eancias) + 2,8202 (rebotes)<\/span><\/p>\n
 Podemos usar o seguinte c\u00f3digo para prever a classifica\u00e7\u00e3o de um novo jogador que tenha 20 pontos, 5 assist\u00eancias e 2 rebotes:<\/span><\/p>\n
 #define new player\n<\/span>new <- data. frame<\/span> (points=c(20), assists=c(5), rebounds=c(2))\n\n#use the fitted model to predict the rating for the new player\n<\/span>predict(model, newdata=new)\n\n       1 \n83.39607 \n<\/strong><\/pre>\n
 O modelo prev\u00ea que este novo jogador ter\u00e1 uma classifica\u00e7\u00e3o de 83,39607<\/strong> .<\/span><\/p>\n
 Podemos confirmar que isso est\u00e1 correto inserindo os valores do novo jogador na equa\u00e7\u00e3o de regress\u00e3o ajustada:<\/span><\/p>\n