4
I’m using a function to calculate regressions. I need the residues of a specific relation to relate to another variable. However I need the waste to be calculated according to Facet grid.
Thus, for each of the divisions the waste will be specific.
This is my code, but it returns null values.
reg = function(data) {
model1 = lm(r ~ a,data= df)
model1_sum = summary(model1)
residuals = as.data.frame(model1$residuals)
df2 = cbind(df,residuals)
names(df2)[names(df2)=="model1$residuals"] <- "residuals"
model2 = lm(residuals ~ a,data= df2)
model2_sum = summary(model2)
formula = sprintf("y= %.3f %+.3f*x",coef(model)[1], coef(model)[2])
r = model2_sum$r.squared
r2 = sprintf("r2= %.3f", r)
x = cor.test(~residuals + lat,data = df2)
r0 = sprintf("r= %.3f", sqrt(model2_sum$r.squared))
p1 = pf(model2_sum$fstatistic[1],model2_sum$fstatistic[2],model2_sum$fstatistic[3],lower.tail=F)
p =sprintf("p = %.3f", p1)
n0 = length(model2_sum$residual)
n1 = sprintf("N = %.f", n0)
data.frame(formula=formula, r=r0,r2=r2, p=p,n=n1, stringsAsFactors=FALSE)
}
df2_math = ddply(data, c("continente","banco"), df2)
df2_math
Joyce, could you post a reproducible example of your data? Also, your code seems to be incomplete: no time do you use the function
reg
, the objectsdf
anddf2
have not been defined and you do not use the function argument within it.– Molx