I have a data frame of individual animals located for different lengths of time. Each row identifies the individual (eg- T003, T121, etc.), the X and Y coordinates in UTMs, and the date the location was collected. I'm trying to calculate average daily distance moved for each individual to create a vector for comparison between individuals/populations. What's the best way to do this in r?
TelemetryID Date Easting Northing
1 T007 9/25/11 739632 3597373
2 T007 8/13/11 739637 3597367
3 T007 8/22/11 739641 3597375
4 T007 9/23/11 739637 3597388
5 T007 8/17/11 739639 3597409
6 T007 9/5/11 739623 3597379
7 T007 8/20/11 739635 3597385
8 T007 9/8/11 739668 3597369
9 T007 8/15/11 739633 3597384
10 T007 9/3/11 739632 3597377
I recognize that these are not consecutive dates, so it requires code function that will recognize calendar date relationships.
The end goal is a vector of average daily distance moved to add as a column to the following data frame
TelemetryID Area Date Sex
1 T001 6.643804e-11 8/10/11 M
2 T002 5.940842e-12 8/7/11 M
3 T003 1.389048e-10 8/10/11 M
4 T004 8.175402e-12 8/7/11 M
5 T005 4.928881e-11 8/9/11 M
6 T006 2.697745e-11 8/10/11 M
7 T007 1.168960e-10 8/10/11 F
Input and Output tables are different because the input table includes every instance of location for an individual, which will by function be distilled to an average value that can be attributed to a single individual; the average value will be a dependent variable in modeling.
result <- SlimBoth %>%
mutate(Date = as.Date(Date, format = "%m/%d/%y")) %>%
arrange(Date) %>%
group_by(TelemetryID) %>%
mutate( Dist = pointDistance(cbind(Easting, Northing),
cbind(lag(Easting), lag(Northing)),
lonlat = FALSE),
Elapsed = as.integer(Date - lag(Date)),
DistPerDay = Dist / Elapsed)
result
result %>%
dplyr::summarise(AveDist = mean(DistPerDay, na.rm = TRUE)) %>%
right_join(Telemetered.1)->ADDM
This function works great, and I updated the telemetered.1 data frame to include the column for Average Daily Distance Moved. The resultant table has a great deal of "Inf" entered where the mean movement values should be.
TelemetryID AveDist Date Easting Northing Sex Translocated
<chr> <dbl> <chr> <int> <int> <chr> <chr>
1 T001 Inf 8/10/11 736408 3598539 M No
2 T002 Inf 8/7/11 736529 3598485 M No
3 T003 Inf 8/10/11 736431 3598671 M No
4 T004 Inf 8/7/11 736535 3598673 M No
5 T005 Inf 8/9/11 739641 3597415 M No
6 T006 30.2 8/10/11 735846 3598974 M No
7 T007 Inf 8/10/11 739647 3597146 F No
8 T008 Inf 8/11/11 739797 3597455 M No
9 T009 Inf 8/11/11 729166 3603726 F No
10 T010 Inf 8/11/11 729058 3603703 M No
The first df includes all of the instances of location for each individual. I want to summarize all these locations per individual with the value Average Daily Distance Moved (ADDM). This will yield 1 value/individual. I then want to add this value to another df for modeling that includes Individual (TelemetryID), sex, translocation status, ADDM, and Area of home range (which I've calculated separately for each individual). Here's data for an individual that was located twice on at least one day:
TelemetryID Date Time Easting Northing Sex Translocated
4969 T237 8/14/13 10:36:00 740968 3597704 M No
4970 T237 8/7/13 10:52:00 740860 3597865 M No
4971 T237 8/13/13 09:49:00 740893 3597835 M No
4972 T237 7/29/13 19:41:00 740872 3597872 M No
4973 T237 8/6/13 10:36:00 741002 3597627 M No
4974 T237 8/17/13 19:13:00 740965 3597710 M No
4975 T237 8/18/13 19:25:00 740964 3597705 M No
4976 T237 8/3/13 10:58:00 740860 3597865 M No
4977 T237 8/5/13 09:20:00 740985 3597695 M No
4978 T237 8/14/13 19:37:00 741005 3597644 M No
4979 T237 7/30/13 10:03:00 740862 3597862 M No
4980 T237 7/31/13 10:37:00 740874 3597862 M No
4981 T237 8/20/13 18:56:00 740916 3597720 M No
4982 T237 8/21/13 05:46:00 741025 3597736 M No
4983 T237 8/27/13 10:07:00 740963 3597828 M No
4984 T237 8/30/13 09:54:00 741019 3597768 M No
4985 T237 9/1/13 11:07:00 740871 3597861 M No
4986 T237 8/28/13 09:51:00 740954 3597626 M No
4987 T237 8/1/13 19:07:00 740880 3597862 M No
One approach would be to use pointDistance from raster and lag from dplyr:
library(dplyr)
library(raster)
result <- data %>%
mutate(DateTime = as.POSIXct(paste(Date,Time), format = "%m/%d/%y %H:%M:%S")) %>%
dplyr::select(TelemetryID, Sex, Translocated, Easting, Northing, DateTime) %>%
arrange(DateTime) %>%
group_by(TelemetryID) %>%
mutate( Dist = pointDistance(cbind(Easting, Northing),
cbind(lag(Easting), lag(Northing)),
lonlat = FALSE),
Elapsed = as.numeric(difftime(DateTime,lag(DateTime),units = "days")),
DistPerDay = Dist / Elapsed)
result
# TelemetryID Sex Translocated Easting Northing DateTime Dist Elapsed DistPerDay
# <fct> <fct> <fct> <int> <int> <dttm> <dbl> <dbl> <dbl>
# 1 T237 M No 740872 3597872 2013-07-29 19:41:00 NA NA NA
# 2 T237 M No 740862 3597862 2013-07-30 10:03:00 14.1 0.599 23.6
# 3 T237 M No 740874 3597862 2013-07-31 10:37:00 12 1.02 11.7
# 4 T237 M No 740880 3597862 2013-08-01 19:07:00 6 1.35 4.43
# 5 T237 M No 740860 3597865 2013-08-03 10:58:00 20.2 1.66 12.2
# 6 T237 M No 740985 3597695 2013-08-05 09:20:00 211. 1.93 109.
# 7 T237 M No 741002 3597627 2013-08-06 10:36:00 70.1 1.05 66.6
# 8 T237 M No 740860 3597865 2013-08-07 10:52:00 277. 1.01 274.
# 9 T237 M No 740893 3597835 2013-08-13 09:49:00 44.6 5.96 7.49
#10 T237 M No 740968 3597704 2013-08-14 10:36:00 151. 1.03 146.
#11 T237 M No 741005 3597644 2013-08-14 19:37:00 70.5 0.376 188.
#12 T237 M No 740965 3597710 2013-08-17 19:13:00 77.2 2.98 25.9
#13 T237 M No 740964 3597705 2013-08-18 19:25:00 5.10 1.01 5.06
#14 T237 M No 740916 3597720 2013-08-20 18:56:00 50.3 1.98 25.4
#15 T237 M No 741025 3597736 2013-08-21 05:46:00 110. 0.451 244.
#16 T237 M No 740963 3597828 2013-08-27 10:07:00 111. 6.18 17.9
#17 T237 M No 740954 3597626 2013-08-28 09:51:00 202. 0.989 204.
#18 T237 M No 741019 3597768 2013-08-30 09:54:00 156. 2.00 78.0
#19 T237 M No 740871 3597861 2013-09-01 11:07:00 175. 2.05 85.2
Now you can summarize the data however you'd like, such as with mean, and join to your other data:
result %>%
summarise(AveDist = mean(DistPerDay, na.rm = TRUE)) %>%
right_join(data2)
## A tibble: 7 x 5
# TelemetryID AveDist Area Date Sex
# <fct> <dbl> <dbl> <fct> <fct>
#1 T237 85.0 6.64e-11 8/10/11 M
#2 T002 NA 5.94e-12 8/7/11 M
#3 T003 NA 1.39e-10 8/10/11 M
#4 T004 NA 8.18e-12 8/7/11 M
#5 T005 NA 4.93e-11 8/9/11 M
#6 T006 NA 2.70e-11 8/10/11 M
#7 T007 NA 1.17e-10 8/10/11 F