I am working with a regression, say:
df <- data.frame(x1=c(1, 2, 4, 5, 5, 6, 6, 7, 8, 10, 11, 11, 12, 12, 14),
x2=c(1, 2, 3, 6, 15, 10, 11, , 29, 12, 11, 3, 34, 27, 4),
y=c(64, 66, 76, 73, 74, 81, 83, 82, 80, 88, 84, 82, 91, 93, 89))
attach(df)
model <- lm(y~x1+x2)
Which looks like:
> summary(model)
Call:
lm(formula = y ~ x1 + x2)
Residuals:
Min 1Q Median 3Q Max
-4.062 -2.478 -1.309 2.787 5.649
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 64.77697 1.96609 32.947 0.000000000000387 ***
x1 1.77351 0.25602 6.927 0.000015886638034 ***
x2 0.17577 0.09692 1.814 0.0948 .
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 3.357 on 12 degrees of freedom
Multiple R-squared: 0.8673, Adjusted R-squared: 0.8452
F-statistic: 39.23 on 2 and 12 DF, p-value: 0.000005451
I then want to use predict(model) to compute prediction values given new data. However, in some cases, it will be meaningful to set some coefficients equal to zero. Here, assume that I want to produce a modified model, say model0 such that:
model0$coefficients
(Intercept) x1 x2
64.7769674 0 0.1757696
My question is if there is a recommended way to do so? I understand this may be equivalent to asking what is the best way to modify the element of a sublist, but would like to know your take.
