Jump to content
Ornithology Exchange


Society Members
  • Posts

  • Joined

  • Last visited

Profile Information

  • Location
  • Country

Recent Profile Visitors

1,019 profile views

AmelieRobertoCharron's Achievements


Newbie (1/14)



  1. Hello! I am currently using FLightR to analyze some geolocator data, and am trying to visualize the error around the migration route that is estimated by the package. I am attempting to plot the mean latitude and longitude from the results with both the 1st and 3rd quartiles and the 95% confidence interval (CI). Essentially, I am trying to merge the figures produced by the map.FLightR.ggmap() function and the plot_lon_lat() function. I have tried plotting the migration route with error bars, as well as plotting the quartiles and confidence intervals as lines in ggplot2. Unfortunately I have found the error bars to be very messy, making it difficult to discern where the mean coordinates are and where the quartiles and CI differ. Additionally, the plotting of the quartiles and CI values as lines is a confusing representation of the uncertainty in the location estimates as they often cross the mean coordinates when the bird switches direction. To attempt to create a range around the migration route that shows the error I also ran an if else function that switched the upper and lower latitude ranges when the bird changes direction to keep the CI or quartile line from crossing the route. Unfortunately, this solution seemed a bit too sensitive and caused more jogs in the upper and lower CI and quantiles than I started with! Would anyone have any recommendations on how to plot the 1st and 3rd quartiles and upper and lower confidence intervals on a migration route? Any recommendations would be appreciated! Thank you, Amélie
  2. Hi Eli, Thanks for your help! Subsetting the data worked. And thanks for your recommendations. I will keep as many transitons as possible. Thanks again, Amélie
  3. I should also add that I have tried removing both sunrise and sunset times for the same day or light curve when there either are unrealistic, and have found that they also aren't deleted when I look at the twl data. Thanks, Amélie
  4. Hi Eli, Thank you for the response. I was trying to delete transitions that using the preprocessLight function in hopes of cleaning up my data to run it in FLightR. I deleted transitions in preprocessLight by clicking the "d" key in the fourth step, and recorded by hand which transitions I deleted, so as to have a reference after to see whether specific transitions were removed. When I looked at the results from the preprocessLight function, none of the transitions I deleted were gone, they were all still present. Below I have included some twilight data, the records in red I attempted to delete and they are still present. Twilight Rise Deleted Marker Inserted Twilight3 Marker3 15/06/2014 5:02 FALSE FALSE 0 FALSE 15/06/2014 5:02 0 15/06/2014 10:21 TRUE TRUE 0 FALSE 15/06/2014 10:21 0 16/06/2014 5:07 FALSE FALSE 0 FALSE 16/06/2014 5:07 0 16/06/2014 10:11 TRUE FALSE 0 FALSE 16/06/2014 10:11 0 17/06/2014 4:52 FALSE FALSE 0 FALSE 17/06/2014 4:52 0 17/06/2014 10:21 TRUE TRUE 0 FALSE 17/06/2014 10:21 0 18/06/2014 4:57 FALSE FALSE 0 FALSE 18/06/2014 4:57 0 18/06/2014 10:01 TRUE FALSE 0 FALSE 18/06/2014 10:01 0 19/06/2014 5:07 FALSE FALSE 0 FALSE 19/06/2014 5:07 0 19/06/2014 10:01 TRUE FALSE 0 FALSE 19/06/2014 10:01 0 20/06/2014 4:47 FALSE TRUE 0 FALSE 20/06/2014 4:47 0 20/06/2014 10:16 TRUE FALSE 0 FALSE 20/06/2014 10:16 0 21/06/2014 5:07 FALSE FALSE 0 FALSE 21/06/2014 5:07 0 21/06/2014 10:16 TRUE FALSE 0 FALSE 21/06/2014 10:16 0 22/06/2014 5:17 FALSE FALSE 0 FALSE 22/06/2014 5:17 0 22/06/2014 10:05 TRUE FALSE 0 FALSE 22/06/2014 10:05 0 23/06/2014 5:12 FALSE FALSE 0 FALSE 23/06/2014 5:12 0 23/06/2014 10:01 TRUE FALSE 0 FALSE 23/06/2014 10:01 0 24/06/2014 5:07 FALSE FALSE 0 FALSE 24/06/2014 5:07 0 24/06/2014 10:05 TRUE FALSE 0 FALSE 24/06/2014 10:05 0 25/06/2014 4:47 FALSE TRUE 0 FALSE 25/06/2014 4:47 0 25/06/2014 10:40 TRUE TRUE 0 FALSE 25/06/2014 10:40 0 26/06/2014 5:07 FALSE FALSE 0 FALSE 26/06/2014 5:07 0 26/06/2014 10:11 TRUE FALSE 0 FALSE 26/06/2014 10:11 0 27/06/2014 4:58 FALSE FALSE 0 FALSE 27/06/2014 4:58 0 27/06/2014 10:11 TRUE FALSE 0 FALSE 27/06/2014 10:11 0 I assume it's possible to delete transitions, as others have reported that they have deleted unrealistic transitions, but I can't seem to make it work in preprocessLight. As an alternative I can export the twilight times and delete them by hand in excel and reimport it into R, but that seems much more complicated than simply doing it in preprocessLight. If you have any advice I'd really appreciate it! Thanks again, Amélie
  5. Hello again, I ran the following code to preprocess with BAStag: # Set threshold threshold = log(4.5) # Preprocess in BAStag twl lmax = 12) In the preprocessLight function I deleted inaccurate twilights with the 'd' key in the four stage which caused them to turn grey. I then moved on and saved my results. # Save result as csv write.csv(twl, paste("twl.csv", sep = "?/"), row.names = F) # Converts twilights into appropriate format twl$Twilight tz= "GMT") # Account for storing maximum values for each 5 minute period. # Adjust twilight data since this tag records maximum values over time (in this # case every 5 minutes, 60*5) twl And to view the results of the preprocessing I ran this: # Show results of defining twilights against raw light data # Offset of 18 puts the night in the middle (so the night isn't cut off) offset = 18 lightImage(d.lux, offset = offset, zlim = c(0,12)) tsimagePoints(twl$Twilight, offset = offset, pch = 16, cex = 0.5, col = ifelse(twl$Rise, "dodgerblue", "firebrick")) When I go to view the results all of the twilights I deleted are still present. Does anyone know how to delete inaccurate twilights? Thanks, Amélie
  6. Hello everyone, How do you delete twilights using the preprocessLight function in BAStag? The help file indicates that at stage four you should be able to use the 'd' to toggle deletions. I have found that this changes the color of the sunrise or sunset from orange or blue to gray, but when I show the results of my defined twilights against the raw data that they appear again. I'm certain it's an easy fix, I'm just unsure how to do it. Thanks for your help! Amélie
  • Create New...