How to get smoothed data as a vector?

[nickcin]

How to get smoothed data as a vector?

For example, I draw a graph

ggplot2::ggplot(data=data.frame(x=seq(1,28),y=data$Var1), aes(x, y)) +
geom_point() +
geom_smooth(method = "loess", se = FALSE, span=0.3, n=28)

however, I don't understand how to extract the result as a vector

[nirgrahamuk]

The best way to get your data smoothed is to use stats::loess function directly rather than have ggplot do it for you

[rene_at_coco]

# package libraries 
library(tidyverse)
#> Warning: package 'tidyverse' was built under R version 4.2.2
#> Warning: package 'ggplot2' was built under R version 4.2.3
#> Warning: package 'tibble' was built under R version 4.2.3
#> Warning: package 'tidyr' was built under R version 4.2.2
#> Warning: package 'readr' was built under R version 4.2.2
#> Warning: package 'purrr' was built under R version 4.2.2
#> Warning: package 'dplyr' was built under R version 4.2.3
#> Warning: package 'stringr' was built under R version 4.2.2
#> Warning: package 'forcats' was built under R version 4.2.2
#> Warning: package 'lubridate' was built under R version 4.2.2

# sample data 
set.seed(9)

df <- tibble(
  x = seq(1, 28, 1),
  var1 = rnorm(n = 28, mean = 1, sd = .1)
)

# apply stats::loess function directly 
df <- df %>%
  mutate(
    m = predict(loess(var1 ~ x))
  )

# view
df
#> # A tibble: 28 × 3
#>        x  var1     m
#>    <dbl> <dbl> <dbl>
#>  1     1 0.923 0.923
#>  2     2 0.918 0.943
#>  3     3 0.986 0.960
#>  4     4 0.972 0.975
#>  5     5 1.04  0.988
#>  6     6 0.881 0.999
#>  7     7 1.12  1.01 
#>  8     8 0.998 1.01 
#>  9     9 0.975 1.02 
#> 10    10 0.964 1.02 
#> # ℹ 18 more rows

^{Created on 2023-08-16 with reprex v2.0.2}

[nickcin]

Thank you very much for your answer! But I think they have different default settings. I liked the result of the "loess" function from "ggplot2" more than from "stats" when I tried to solve the problem. I'll try again as you suggested.

[nickcin]

Thank you very much for your answer! Everything is clear, now I'll try, similar code!

ps. Is it really impossible to get a result from ggplot2? I rummaged through the structure of the returned object, but found nothing.

[nickcin]

The default result of stats::loess is really bad. How to remove the shift and improve anti-aliasing?

x<-seq(1,28)
var1<-data$Trechus.secalis
var1
 [1]  0  0  0  1  1  5  2  8  5  5  0  3  1  0  3  5  1  2  1  0  2 12  0  2  3  0
[27]  2  0

plot(x,var1,pch=19,cex=1)
j <- order(var1)
l<-loess(var1~x)
lines(x,l$fitted[j],col="red",lwd=3)

[nickcin]

l<-predict(loess(var1 ~ x,span=0.3))
lines(x,l,col="cyan",lwd=3)