Hi everyone !
I am looking to obtain this type of graph where each trajectory/line represents a smoothed average of all the lines within a given category (either "w" or "l" in my case). Additionally, I would like to include standard error (in a more transparent color on the figure) on each line.
Here is a sample of my data
The values of y given in columns F10 to F110 correspond to a single data point on each trajectory/line.
I would be very grateful for any help :))
Thanks in advance.
Hello @PhonPhon and welcome to Posit Community.
It would help if you could share your example dataset in a way that is easy for your potential helpers to use. I suggest using the dput() function on your df variable and sharing the output here:
dput(df)
Ok, sure ! Here is a sample subset of my data:
dput(df)
structure(list(syllabe = c("CV", "CV", "VCs", "CV", "CV", "CVs",
"CV", "CV", "CV", "CV"), C = c("w", "w", "l", "w", "w", "l",
"w", "l", "w", "l"), V = c("e", "e", "a", "o", "e", "e", "e",
"e", "e", "e"), F10 = c(1276L, 650L, 637L, 626L, 591L, 1868L,
595L, 555L, 675L, 1787L), F11 = c(1088L, 641L, 633L, 602L, 557L,
1846L, 595L, 581L, 670L, 1932L), F12 = c(568L, 628L, 627L, 578L,
539L, 1825L, 598L, 530L, 665L, 1745L), F13 = c(571L, 604L, 619L,
608L, 528L, 1799L, 542L, 487L, 663L, 1814L), F14 = c(587L, 565L,
610L, 691L, 536L, 1774L, 486L, 490L, 660L, 1768L), F15 = c(660L,
523L, 595L, 715L, 542L, 1190L, 490L, 566L, 658L, 1735L), F16 = c(657L,
503L, 579L, 699L, 547L, 443L, 496L, 589L, 650L, 1715L), F17 = c(558L,
515L, 564L, 650L, 562L, 472L, 489L, 1797L, 641L, 1694L), F18 = c(530L,
547L, 547L, 610L, 584L, 477L, 493L, 1802L, 635L, 1687L), F19 = c(500L,
575L, 525L, 575L, 597L, 499L, 503L, 494L, 629L, 1690L), F110 = c(1771L,
532L, 507L, 585L, 572L, 641L, 530L, 580L, 492L, 504L)), row.names = c(NA,
10L), class = "data.frame")
you don't appear to have the numbers that would be suitable for x axis.
interval ?
Actually, since there are 11 columns (corresponding to measurements taken at as many time intervals), I would like the x axis to correspond to these intervals
library(tidyverse)
start_data <- tibble::tribble(
~syllabe, ~C, ~V, ~F10, ~F11, ~F12, ~F13, ~F14, ~F15, ~F16, ~F17, ~F18, ~F19, ~F110,
"CV", "w", "e", 1276L, 1088L, 568L, 571L, 587L, 660L, 657L, 558L, 530L, 500L, 1771L,
"CV", "w", "e", 650L, 641L, 628L, 604L, 565L, 523L, 503L, 515L, 547L, 575L, 532L,
"VCs", "l", "a", 637L, 633L, 627L, 619L, 610L, 595L, 579L, 564L, 547L, 525L, 507L,
"CV", "w", "o", 626L, 602L, 578L, 608L, 691L, 715L, 699L, 650L, 610L, 575L, 585L,
"CV", "w", "e", 591L, 557L, 539L, 528L, 536L, 542L, 547L, 562L, 584L, 597L, 572L,
"CVs", "l", "e", 1868L, 1846L, 1825L, 1799L, 1774L, 1190L, 443L, 472L, 477L, 499L, 641L,
"CV", "w", "e", 595L, 595L, 598L, 542L, 486L, 490L, 496L, 489L, 493L, 503L, 530L,
"CV", "l", "e", 555L, 581L, 530L, 487L, 490L, 566L, 589L, 1797L, 1802L, 494L, 580L,
"CV", "w", "e", 675L, 670L, 665L, 663L, 660L, 658L, 650L, 641L, 635L, 629L, 492L,
"CV", "l", "e", 1787L, 1932L, 1745L, 1814L, 1768L, 1735L, 1715L, 1694L, 1687L, 1690L, 504L
)
(long_data <- start_data |> select(-syllabe ,-V) |> pivot_longer(cols = -C) )
(smry_data <- long_data |> group_by(C,name) |>
summarise(avg=mean(value),
stderr=sd(value)/sqrt(n()),
upper=avg+stderr,
lower=avg-stderr))
ggplot(data=smry_data) + aes(
x=name,
y=avg,color=C,
group=C
) + geom_line()+
geom_ribbon(mapping = aes(ymin=lower,
ymax=upper),alpha=0.1,color=rgb(0,0,0,0))
Thank you for your response! Excuse my beginner question, but I am struggling to convert my data.frame (with a few thousand rows) to the tibble (or tribble) format. I tried using datapasta::dpasta(), but I am stuck at this point.
By the way, is it possible to add a curve smoothing on the graph?
dat2 <- as_tibble(dat1)
You dont need a tribble. A tribble is a way to turn text into a data.frame. but you start out with a data.frame. so just assume that step.
This topic was automatically closed 42 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.