Having completed my regression equation and F-statistic I have been able to plot a regression curve with confidence limits through my set of data points which I am happy with. My x variable is number of days from start of observations which runs for exactly one year (17 Jan 2009 to 16 Jan 2010). The resulting plot has an x-axis with ticks at 50 day intervals with scale marked at 100 day intervals in figures.
I would like to change the tick intervals to days 15,74,135,196,257 and 318 and marked 1 Feb, 1 Apr, 1 Jun, 1 Aug, 1 Oct and 1 Dec.
**This is the coding I used for the regression and original plot
frpd<-read_csv("frpd.csv")
ggplot(frpd, aes(x=Days, y=cover))+geom_point()
frpd.shuffled<-frpd[sample(nrow(frpd)),]
K<-10
degree<-5
folds<-cut(seq(1,nrow(frpd.shuffled)),breaks=K, labels=FALSE)
mse=matrix(data=NA,nrow=K,ncol=degree)
for(i in 1:K){
testIndexes <- which(folds==i,arr.ind=TRUE)
testData <- frpd.shuffled[testIndexes, ]
trainData <- frpd.shuffled[-testIndexes, ]
for (j in 1:degree){
fit.train = lm(cover ~ poly(Days,j), data=trainData)
fit.test = predict(fit.train, newdata=testData)
mse[i,j] = mean((fit.test-testData$cover)^2)
}
}
colMeans(mse)
best = lm(cover ~ poly(Days,2, raw=T), data=frpd)
summary(best)
lm(formula=cover~poly(Days,2,raw=T),data=frpd)
ggplot(frus, aes(x=Days, y=cover)) +
geom_point() +
stat_smooth(method='lm', formula = y ~ poly(x,2), size = 1) +
xlab('Days') +
ylab('cover')
I changed the plot code to the following but pretty much by guess work:
ggplot(frpd, aes(x=Days, y=cover))
geom_point() +
stat_smooth(method='lm', formula = y ~ poly(x,2), size = 1) +
geom_point(mapping = frpd,x=c(15,74,135,196,257,318),y=cover, xaxt='n')+
ylab('cover')+
xlab('date')+
plot(axis(1,at=c(15,74,135,196,257,318)),labels=c("1 Feb", "1 Apr", "1 Jun", "1 Aug", "1 Oct", "1 Dec"))
scale_y_continuous(limits = c(0,100), breaks=seq(0,100,20))
When I ran this I got the following error report:
> ggplot(frpd, aes(x=Days**, y=cover))
> geom_point() +
+ stat_smooth(method='lm', formula = y ~ poly(x,2), size = 1) +
+ geom_point(mapping = frpd,x=c(15,74,135,196,257,318),y=cover, xaxt='n')+
+ ylab('cover')+
+ xlab('date')+
+ plot(axis(1,at=c(15,74,135,196,257,318)),labels=c("1 Feb", "1 Apr", "1 Jun", "1 Aug", "1 Oct", "1 Dec"))
Error in `+.gg`:
! Cannot add <ggproto> objects together
ℹ Did you forget to add this object to a <ggplot> object?
Run `rlang::last_trace()` to see where the error occurred.
> scale_y_continuous(limits = c(0,100), breaks=seq(0,100,20))
<ScaleContinuousPosition>
Range:
Limits: 0 -- 100
**My data starts like this: **
> frpd
# A tibble: 72 × 3
Date Days cover
<dbl> <dbl> <dbl>
1 39863 49 100
2 39885 71 100
3 39966 152 50
4 40065 251 13
5 40109 295 25
6 40151 337 25
7 40196 382 25
8 39870 56 100
9 39885 71 100
10 39935 121 17
I would certainly appreciate help
I started with the code I found in R specifying x-axis ticks for a line plot
But could not figure how to adapt it from a linear plot of points to smoothed output of the polynomial regression.
I'm not sure what else to try.
Images of plots: 1 Original with full confidence limits overlay
i think the plot() function is what causes the error. you could use scale_x_continuous() to fix that like this:
ggplot(frpd, aes(x = Days, y = cover)) +
geom_point() +
stat_smooth(method = 'lm', formula = y ~ poly(x, 2), size = 1) +
scale_x_continuous(breaks = c(15, 74, 135, 196, 257, 318),
labels = c("1 Feb", "1 Apr", "1 Jun", "1 Aug", "1 Oct", "1 Dec")) +
scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, 20)) +
ylab('cover') +
xlab('date')