ax3man / trackr Goto Github PK

For a large number of trials, with many mistakes, this function takes a long time. Have a look at data.table::foverlaps.

Hi,
I downloaded your package and as I was going through it using your example data I came across this error:

Error in get(as.character(FUN), mode = "function", envir = envir) :
object 'turn' of mode 'function' was not found

I have installed all the mentioned packages with the library function, and followed every single step you posted, also tried doing it all with my data but it still shows the same error, could you help me?

Thank you!
Ari

Make it work with soc as well, preferably auto detect location of variables (use tracks$pr).

It should recognize when .dots are not function calls, and ideally should also fill in defaults for more complex expressions than single function calls.

Error occurs in the last line of the function.

In some cases it appears that reading in single files or many files gives different outputs for some trials. Might be related only to plotting. Check Simons email.

[, [[ and $ should instead be used to select tables.

add end_new = ~if_else(r == n(), end, end_new) to the mutate call.

The package is automatically combining data from different trials into one trial. I have realized that the way that I have named the files is causing this problem. For instance, I have several individuals from the same treatment which I have named XXXX1.1.csv; XXXX1.2.csv; XXXX1.3.csv, and the tracks from these individuals are then combined in the data as only one track. I have renamed the files using a numerical sequence and this problem disappear.

I attach an example with 9 files that get combined into only 6 tracks using the following code:

files<-list.files(startingDir, pattern="csv")
dat <- read_Ctrax(files, type = 'fixed')
tr <- as_tracks(dat, frame_rate = 30)
plot(tr)

tracktest.zip

Thanks!

The error appears when executing the code:
mutate(tr, speed = speed(), acc = acceleration(), turn = angular_velocity())

Manual arena's should be allowed, as well as arenas of fixed size but unknown location.

Eventually some simple tools to extract arena locations from screenshots.

I have tried to read a large amount of Ctrax .csv files into tracker but I get an error message:

Error in names(x) <- do.call(paste0, as.list(new.names)) :
'names' attribute [6] must be the same length as the vector [1]

I can read multiple files using the same procedure with a small subset of these same files. Therefore, I assume that some of the .csv files that I provide must be incorrect, but currently is not possible to know which one is wrong without going back to the .csv files one by one. Is it possible to add that information to the error message?

Thanks!

Error: unknown column 'var'
Check if some or all have to missing for the error.

Solve by checking which vars are present. Warn if some are missing (and don't display), give descriptive error when all are missing.

Currently broken since info is not retrieved from party_df.

Reported statistics seem fine, but rows are duplicated for each animal. Grouping is only by trial, which is correct. Source of confusion unclear.

$animal is not grouped?

tr <- estimate_arena(tr, grouping = NULL, shape = "rectangle",height = 1, width = 2) fails for the guppies example.

Grouping with "rectangle" fails.