CODEWITHSUNDEEP
rggplot2curve-fitting
Agus camacho
Agus camacho

Reputation: 898

apparently correct rscript does not plot curve in ggplot 2

I have tried different script options from other related questions but none seems to plot the desired curve, despite the script looks correct and runs... https://www.dropbox.com/s/efxbx6qh4z32snx/examplecurve.csv?dl=0

d=read.csv("examplecurve.csv")
require(nlme)
model <- nls(y ~ a*(x^b),start = list(a = exp(coef(m)[1]), b = 
coef(m)[2]),  data = d)
summary(model)
fontsize=12
ggplot(d, aes(x=x, y=y)) +
geom_point(col="black", fill = "grey",alpha=0.8, shape=21, size = 2) +
stat_smooth(method = "nls", 
        method.args = list(formula = y ~ a*(x^b),
    method.args = list(formula = y ~ a*(x^b), 
    start = list(coef(model))),
    se = F, size = 0.5, data = d))+
                        
   theme(panel.background = element_rect(fill='transparent', 
 colour='black'),
    panel.grid.major = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title.x = element_text(colour="black", size=fontsize),
    axis.text.x  = element_text(angle=0, vjust=0.5,colour="black", 
 size=fontsize),
    axis.title.y = element_text(colour="black", size=fontsize),
    axis.text.y  = element_text(angle=0, vjust=0.5,colour="black", 
 size=fontsize)
 )+  theme(legend.position="none",
        legend.justification=c(1,1),
        legend.key = element_rect(fill = 'transparent'),
        legend.background = element_rect(fill = 'transparent'),
        legend.text = element_text(size = 12))

I would also like to add error ribbons...

T hanks in advance!

Upvotes: 1

Views: 67

Answers (2)

Allan Cameron
Allan Cameron

Reputation: 173793

Borrowing from this previous answer of mine, you can add an error ribbon around an nls fit using these little functions:

nls_se <- function(formula, data, start, ...) {
  mod <- nls(formula, data, start)
  class(mod) <- "nls_se"
  mod
}

predict.nls_se <- function(model, newdata, level = 0.9, ...) {
  class(model) <- "nls"
  p <- investr::predFit(model, newdata = newdata, 
                        interval = "confidence", level = level)
  list(fit = p, se.fit = p[,3] - p[,1])
}

These allow you to use geom_smooth directly with your target formula:

ggplot(d, aes(x = x, y = y)) +
  geom_point(col="black", fill = "grey",alpha = 0.8, shape = 21, size = 2) +
  geom_smooth(method = nls_se, formula = y ~ a * x^b,
              method.args = list(start = list(a = 1, b = 1)))

enter image description here

Or, with some styling choices (thanks Rui Barradas)

ggplot(d, aes(x = x, y = y)) +
  geom_point(col="black", fill = "grey",alpha = 0.8, shape = 21, size = 2) +
  geom_smooth(method = nls_se, formula = y ~ a * x^b,
              method.args = list(start = list(a = 1, b = 1)),
              color = "navy", fill = "deepskyblue4", alpha = 0.2,
              linetype = 2, linewidth = 0.4) +
  theme_Agus_camacho(20)

enter image description here

Upvotes: 2

Rui Barradas
Rui Barradas

Reputation: 76402

First of all, there is no model m created in the question so I fitted a OLS model to have the rest of the code work. As it turned out, it wasn't a bad choice.

Don't put the formula in the method.args list, formula is a stat_smooth argument and it's better to have it where it is expected to be.

I have created a custom theme with your settings (code at the end) in order not to distract from the main problem. Note that fontsize is passed as an argument to the custom theme.

library(nlme)
library(ggplot2)

fontsize <- 12

m <- lm(y ~ x, d)
model <- nls(y ~ a*(x^b),start = list(a = exp(coef(m)[1]), 
                                      b = coef(m)[2]),  data = d)

ggplot(d, aes(x=x, y=y)) +
  geom_point(col="black", fill = "grey",alpha=0.8, shape=21, size = 2) +
  stat_smooth(method = "nls", 
              formula = y ~ a * x^b,
              method.args = list(start = list(a = exp(coef(m)[1]), 
                                              b = coef(m)[2])),
              se = FALSE, linewidth = 0.5) +
  theme_Agus_camacho(fontsize)

Created on 2022-12-12 with reprex v2.0.2

Custom theme

theme_Agus_camacho <- function(fontsize){ 
  theme_minimal(base_size = fontsize) %+replace%
    theme(
      panel.background = element_rect(fill='transparent', 
                                      colour='black'),
      panel.grid.major = element_blank(),
      panel.grid.minor = element_blank(),
      axis.title.x = element_text(colour="black", size=fontsize),
      axis.text.x  = element_text(angle=0, vjust=0.5,colour="black", 
                                  size=fontsize),
      axis.title.y = element_text(colour="black", size=fontsize),
      axis.text.y  = element_text(angle=0, vjust=0.5,colour="black", 
                                  size = fontsize),
      legend.position="none",
      legend.justification=c(1,1),
      legend.key = element_rect(fill = 'transparent'),
      legend.background = element_rect(fill = 'transparent'),
      legend.text = element_text(size = fontsize)
    )
}

Created on 2022-12-12 with reprex v2.0.2

Upvotes: 2

Related Questions