Goal: to load a movement study from Movebank, obtain the ancillary individual ID data, download additional individual specific data, and link to the movebank data. We also cover some basic plotting and mapping of the movebank data. We will work with data from the Inuvik woodland population (follow link to open the study page on Movebank).

1 Loading data from Movebank

From within R, load the move package:

require(move)

Create a login object, using your own movebank.org login and password, as follows:

mylogin <- movebankLogin(username="MyUserName", password="MyPassword")

To load a study, it is important that it be spelled out correctly, exactly as it is called in the movebank page. The following command will load the Inuvik data:

inuvik <- getMovebankData("ABoVE: NWT Inuvik Boreal Woodland Caribou", login = mylogin)
## Error in curl::curl_fetch_memory(url, handle = handle): Could not resolve host: www.movebank.org

This step might take some time. But when it is done you will have a new object, called inuvik in yuor workspace. This is the entirety of that study’s data. You will want to save this object locally, as an R object, so you can load it immediately as needed:

save(inuvik, file = "./_data/inuvik.rda")

2 Structure of movebank data

The structure of the inuvik object is rather complex. Entering inuvik into the console will spit out a lot of information:

inuvik
## class       : MoveStack 
## features    : 69997 
## extent      : -136.024, -129.7156, 66.134, 68.8489  (xmin, xmax, ymin, ymax)
## coord. ref. : +proj=longlat +ellps=WGS84 +datum=WGS84 +towgs84=0,0,0 
## variables   : 15
## names       : sensor_type_id, comments, habitat, import_marked_outlier, location_lat, location_long, manually_marked_outlier,           timestamp,               update_ts, argos_calcul_freq, argos_lc, deployment_id,   event_id,    tag_id,         sensor_type 
## min values  :            653,       NA,      NA,                    NA,      66.1340,     -129.7156,                   false, 2002-05-02 00:00:00, 2018-09-17 20:32:20.291,         401648030,        1,     532845467, 7411868493, 532618681, Argos Doppler Shift 
## max values  :          82798,       NA,      NA,                    NA,      68.8489,     -136.0240,                   false, 2012-07-05 16:01:55, 2018-09-17 20:35:37.738,         401652084,        3,     532845511, 7411957236, 532839235,                 GPS 
## timestamps  : 2002-05-02 00:00:00 ... 2012-07-05 16:01:55 Time difference of 3718 days  (start ... end, duration) 
## sensors     : GPS, Argos Doppler Shift 
## indiv. data : comments, death_comments, earliest_date_born, exact_date_of_birth, individual_id, latest_date_born, local_identifier, nick_name, ring_id, sex, taxon_canonical_name 
## min ID Data : NA, NA, NA, NA, 532618680, NA,     BW26, IN0701, NA, f, caribou 
## max ID Data : NA, NA, NA, NA, 532622218, NA, BWC36189,   IN45, NA, f, caribou 
## individuals : BW26, BW27, BW28, BW29, BW30, BW31, BW32, BW33, BW34, BW35, BW36, BW37, BW39, BW40, BW41 
## unused rec. : 243 
## license     : Data are for use only as part of NASA's Arctic-Boreal Vul ...
## citation    : Nagy JA, Derocher AE, Nielsen SE, Wright WH, Heikkila JM (2006) Modelling seasonal habitats of boreal woodland caribou at the northern limits of their range: a preliminary assessment of the Lower Mackenzie River Valley, Northwest Territories, Canada. Government of the Northwest Territories. <br><br> Nagy JA, Auriat D, Wright W, Slack T, Ellsworth I, Kienzler M (2004) Ecology of boreal woodland caribou in the Lower Mackenzie Valley NT: work completed in the Inuvik region April 2003 to November 2004. <br><br> See others at www.enr.gov.nt.ca/resources. 
## study name  : ABoVE: NWT Inuvik Boreal Woodland Caribou 
## date created: 2018-09-14 21:44:18

In fact, this is a movestack object, which contains information inside of elements called slots. Eg: a dataframe has column, but a movestack is an S4 “formal” dataclass which has strictly defined slots. This data structure is typical for spatial data.

2.1 Slots

To see the names of the slots:

slotNames(inuvik)
##  [1] "trackId"                 "timestamps"             
##  [3] "idData"                  "sensor"                 
##  [5] "data"                    "coords.nrs"             
##  [7] "coords"                  "bbox"                   
##  [9] "proj4string"             "trackIdUnUsedRecords"   
## [11] "timestampsUnUsedRecords" "sensorUnUsedRecords"    
## [13] "dataUnUsedRecords"       "dateCreation"           
## [15] "study"                   "citation"               
## [17] "license"

You can access slots with the @ tag (contrast to the $ for data frames and lists). The data slot contains the movement data:

head(inuvik@data)
sensor_type_id comments habitat import_marked_outlier location_lat location_long manually_marked_outlier timestamp update_ts argos_calcul_freq argos_lc deployment_id event_id tag_id sensor_type
25915 653 NA NA NA 67.5443 -132.1738 false 2005-03-11 00:00:00 2018-09-17 20:35:37.738 NA NA 532845469 7411907216 532618725 GPS
25916 653 NA NA NA 67.5447 -132.1735 false 2005-03-11 08:03:42 2018-09-17 20:35:37.738 NA NA 532845469 7411907218 532618725 GPS
25917 653 NA NA NA 67.5448 -132.1736 false 2005-03-11 16:01:55 2018-09-17 20:35:37.738 NA NA 532845469 7411907217 532618725 GPS
25918 653 NA NA NA 67.5434 -132.1618 false 2005-03-12 00:00:00 2018-09-17 20:35:37.738 NA NA 532845469 7411907219 532618725 GPS
25919 653 NA NA NA 67.5432 -132.1503 false 2005-03-12 07:58:05 2018-09-17 20:35:37.738 NA NA 532845469 7411907221 532618725 GPS
25920 653 NA NA NA 67.5432 -132.1504 false 2005-03-12 16:01:55 2018-09-17 20:35:37.738 NA NA 532845469 7411907220 532618725 GPS

Other slots contain, e.g., the projection, the license, a citation:

inuvik@proj4string
## CRS arguments:
##  +proj=longlat +ellps=WGS84 +datum=WGS84 +towgs84=0,0,0
inuvik@license
## [1] "Data are for use only as part of NASA's Arctic-Boreal Vulnerability Experiment (ABoVE, limited to select ABoVE researchers) and by NWT staff."
inuvik@citation
## [1] "Nagy JA, Derocher AE, Nielsen SE, Wright WH, Heikkila JM (2006) Modelling seasonal habitats of boreal woodland caribou at the northern limits of their range: a preliminary assessment of the Lower Mackenzie River Valley, Northwest Territories, Canada. Government of the Northwest Territories. <br><br> Nagy JA, Auriat D, Wright W, Slack T, Ellsworth I, Kienzler M (2004) Ecology of boreal woodland caribou in the Lower Mackenzie Valley NT: work completed in the Inuvik region April 2003 to November 2004. <br><br> See others at www.enr.gov.nt.ca/resources."

2.2 Individual ID information

The individual ID data is found in a “slot” called idData, with one row per individual:

inuvik@idData
comments death_comments earliest_date_born exact_date_of_birth individual_id latest_date_born local_identifier nick_name ring_id sex taxon_canonical_name
BW26 NA NA NA NA 532618724 NA BW26 IN26 NA f caribou
BW27 NA NA NA NA 532618742 NA BW27 IN27 NA f caribou
BW28 NA NA NA NA 532618733 NA BW28 IN28 NA f caribou
BW29 NA NA NA NA 532618736 NA BW29 IN29 NA f caribou
BW30 NA NA NA NA 532618739 NA BW30 IN30 NA f caribou
BW31 NA NA NA NA 532618729 NA BW31 IN31 NA f caribou
BW32 NA NA NA NA 532622170 NA BW32 IN32 NA f caribou
BW33 NA NA NA NA 532622185 NA BW33 IN33 NA f caribou
BW34 NA NA NA NA 532618746 NA BW34 IN34 NA f caribou
BW35 NA NA NA NA 532622188 NA BW35 IN35 NA f caribou
BW36 NA NA NA NA 532622191 NA BW36 IN36 NA f caribou
BW37 NA NA NA NA 532622194 NA BW37 IN37 NA f caribou
BW39 NA NA NA NA 532622197 NA BW39 IN39 NA f caribou
BW40 NA NA NA NA 532622200 NA BW40 IN40 NA f caribou
BW41 NA NA NA NA 532622203 NA BW41 IN41 NA f caribou
BW42 NA NA NA NA 532622206 NA BW42 IN42 NA f caribou
BW43 NA NA NA NA 532618749 NA BW43 IN43 NA f caribou
BW44 NA NA NA NA 532618799 NA BW44 IN44 NA f caribou
BW45 NA NA NA NA 532618765 NA BW45 IN45 NA f caribou
BWC07.01 NA NA NA NA 532618770 NA BWC07/01 IN0701 NA f caribou
BWC07.04 NA NA NA NA 532618773 NA BWC07/04 IN0704 NA f caribou
BWC07.05 NA NA NA NA 532618784 NA BWC07/05 IN0705 NA f caribou
BWC07.09 NA NA NA NA 532618790 NA BWC07/09 IN0709 NA f caribou
BWC08.01 NA NA NA NA 532618680 NA BWC08/01 IN0801 NA f caribou
BWC08.02 NA NA NA NA 532618690 NA BWC08/02 IN0802 NA f caribou
BWC08.05 NA NA NA NA 532618693 NA BWC08/05 IN0805 NA f caribou
BWC08.07 NA NA NA NA 532618752 NA BWC08/07 IN0807 NA f caribou
BWC08.08 NA NA NA NA 532618696 NA BWC08/08 IN0808 NA f caribou
BWC08.09 NA NA NA NA 532618700 NA BWC08/09 IN0809 NA f caribou
BWC08.10 NA NA NA NA 532618703 NA BWC08/10 IN0810 NA f caribou
BWC08.11 NA NA NA NA 532618706 NA BWC08/11 IN0811 NA f caribou
BWC18 NA NA NA NA 532618755 NA BWC18 IN18 NA f caribou
BWC21 NA NA NA NA 532618709 NA BWC21 IN21 NA f caribou
BWC23 NA NA NA NA 532618758 NA BWC23 IN23 NA f caribou
BWC35981 NA NA NA NA 532618761 NA BWC35981 IN35981 NA f caribou
BWC35982 NA NA NA NA 532618713 NA BWC35982 IN35982 NA f caribou
BWC35983 NA NA NA NA 532618762 NA BWC35983 IN35983 NA f caribou
BWC35984 NA NA NA NA 532618721 NA BWC35984 IN35984 NA f caribou
BWC36182 NA NA NA NA 532618764 NA BWC36182 IN36182 NA f caribou
BWC36186 NA NA NA NA 532622209 NA BWC36186 IN36186 NA f caribou
BWC36187 NA NA NA NA 532622214 NA BWC36187 IN36187 NA f caribou
BWC36188 NA NA NA NA 532622215 NA BWC36188 IN36188 NA f caribou
BWC36189 NA NA NA NA 532622218 NA BWC36189 IN36189 NA f caribou

Unfortunately, none of the columns in the @idData data frame actually corresponds to a column in the @data frame. However, the local_identifier column DOES correspond to another slot in the move object: the trackId:

table(inuvik@trackId)
## 
##     BW26     BW27     BW28     BW29     BW30     BW31     BW32     BW33 
##     4471     2122     2138     2101     2111     2127        5      229 
##     BW34     BW35     BW36     BW37     BW39     BW40     BW41     BW42 
##     5519      139      139      288      509      732      456      565 
##     BW43     BW44     BW45 BWC07.01 BWC07.04 BWC07.05 BWC07.09 BWC08.01 
##       15        4     3608     3527     2470     3604     1145     3208 
## BWC08.02 BWC08.05 BWC08.07 BWC08.08 BWC08.09 BWC08.10 BWC08.11    BWC18 
##     1265     1619      442     3519     2907     3231     3028      116 
##    BWC21    BWC23 BWC35981 BWC35982 BWC35983 BWC35984 BWC36182 BWC36186 
##     4051      319     1303     1562     1164     1365     1996      275 
## BWC36187 BWC36188 BWC36189 
##       70      257      276

This is a strange feature in move objects which I just don’t understand … why not have a consistently names column in both data frames?

2.3 Converting to a dataframe

The move package uses the trackId to knows which animal is which. When you convert the move object to a data frame (using the as.data.frame command), all of the information from the the id table carries over:

inuvik.df <- as.data.frame(inuvik)
names(inuvik.df)
##  [1] "sensor_type_id"          "comments"               
##  [3] "habitat"                 "import_marked_outlier"  
##  [5] "location_lat"            "location_long"          
##  [7] "manually_marked_outlier" "timestamp"              
##  [9] "update_ts"               "argos_calcul_freq"      
## [11] "argos_lc"                "deployment_id"          
## [13] "event_id"                "tag_id"                 
## [15] "sensor_type"             "location_long.1"        
## [17] "location_lat.1"          "optional"               
## [19] "sensor"                  "timestamps"             
## [21] "trackId"                 "comments.1"             
## [23] "death_comments"          "earliest_date_born"     
## [25] "exact_date_of_birth"     "individual_id"          
## [27] "latest_date_born"        "local_identifier"       
## [29] "nick_name"               "ring_id"                
## [31] "sex"                     "taxon_canonical_name"

Note that now we have the trackId, the tag_id, and lots of other ID’s, so we can cross-reference if needed:

inuvik.df[,c("trackId", "local_identifier", "nick_name", "deployment_id", "tag_id","individual_id")] %>% unique
##        trackId local_identifier nick_name deployment_id    tag_id
## 25915     BW26             BW26      IN26     532845469 532618725
## 28068     BW26             BW26      IN26     532845476 532618727
## 38884     BW27             BW27      IN27     532845495 532618743
## 32523     BW28             BW28      IN28     532845510 532618734
## 34663     BW29             BW29      IN29     532845472 532618737
## 36768     BW30             BW30      IN30     532845483 532618740
## 30390     BW31             BW31      IN31     532845474 532618730
## 66339     BW32             BW32      IN32     532845505 532622171
## 66345     BW33             BW33      IN33     532845499 532622186
## 41014     BW34             BW34      IN34     532845473 532618747
## 43005     BW34             BW34      IN34     532845485 532618767
## 66574     BW35             BW35      IN35     532845486 532622189
## 66713     BW36             BW36      IN36     532845479 532622192
## 66852     BW37             BW37      IN37     532845506 532622195
## 67140     BW39             BW39      IN39     532845471 532622198
## 67649     BW40             BW40      IN40     532845496 532622201
## 68382     BW41             BW41      IN41     532845484 532622204
## 68838     BW42             BW42      IN42     532845468 532622207
## 46547     BW43             BW43      IN43     532845509 532618750
## 66335     BW44             BW44      IN44     532845501 532618800
## 51917     BW45             BW45      IN45     532845487 532839235
## 55525 BWC07.01         BWC07/01    IN0701     532845497 532618771
## 59081 BWC07.04         BWC07/04    IN0704     532845478 532618774
## 61566 BWC07.05         BWC07/05    IN0705     532845467 532618785
## 65185 BWC07.09         BWC07/09    IN0709     532845480 532618791
## 2     BWC08.01         BWC08/01    IN0801     532845511 532618681
## 3221  BWC08.02         BWC08/02    IN0802     532845491 532618691
## 4495  BWC08.05         BWC08/05    IN0805     532845470 532618694
## 46566 BWC08.07         BWC08/07    IN0807     532845477 532618753
## 6126  BWC08.08         BWC08/08    IN0808     532845488 532618697
## 9660  BWC08.09         BWC08/09    IN0809     532845503 532618701
## 12578 BWC08.10         BWC08/10    IN0810     532845507 532618704
## 15826 BWC08.11         BWC08/11    IN0811     532845490 532618707
## 47011    BWC18            BWC18      IN18     532845475 532618756
## 18875    BWC21            BWC21      IN21     532845492 532618710
## 47129    BWC23            BWC23      IN23     532845504 532618759
## 47448 BWC35981         BWC35981   IN35981     532845493 532618743
## 22955 BWC35982         BWC35982   IN35982     532845489 532618714
## 48754 BWC35983         BWC35983   IN35983     532845508 532618730
## 24533 BWC35984         BWC35984   IN35984     532845482 532618722
## 49918 BWC36182         BWC36182   IN36182     532845502 532618747
## 69403 BWC36186         BWC36186   IN36186     532845498 532622210
## 69678 BWC36187         BWC36187   IN36187     532845500 532622192
## 69748 BWC36188         BWC36188   IN36188     532845494 532622216
## 70005 BWC36189         BWC36189   IN36189     532845481 532622219
##       individual_id
## 25915     532618724
## 28068     532618724
## 38884     532618742
## 32523     532618733
## 34663     532618736
## 36768     532618739
## 30390     532618729
## 66339     532622170
## 66345     532622185
## 41014     532618746
## 43005     532618746
## 66574     532622188
## 66713     532622191
## 66852     532622194
## 67140     532622197
## 67649     532622200
## 68382     532622203
## 68838     532622206
## 46547     532618749
## 66335     532618799
## 51917     532618765
## 55525     532618770
## 59081     532618773
## 61566     532618784
## 65185     532618790
## 2         532618680
## 3221      532618690
## 4495      532618693
## 46566     532618752
## 6126      532618696
## 9660      532618700
## 12578     532618703
## 15826     532618706
## 47011     532618755
## 18875     532618709
## 47129     532618758
## 47448     532618761
## 22955     532618713
## 48754     532618762
## 24533     532618721
## 49918     532618764
## 69403     532622209
## 69678     532622214
## 69748     532622215
## 70005     532622218

Note that animal BW26 has, for example, multiple deployments.

3 Merging with external ID data

3.1 Ancillary data

For this population (and, apparently, many of the NWT woodland populations), there is some individual specific ancillary data avaialble in a an external file:

inuvik.ref.data <- read.csv("./_data/ABoVE_ NWT Inuvik Boreal Woodland Caribou-reference-data.csv")
names(inuvik.ref.data)
##  [1] "ï..tag.id"               "animal.id"              
##  [3] "animal.taxon"            "deploy.on.date"         
##  [5] "deploy.off.date"         "animal.sex"             
##  [7] "animal.taxon.detail"     "deploy.off.latitude"    
##  [9] "deploy.off.longitude"    "deploy.on.latitude"     
## [11] "deploy.on.longitude"     "deployment.end.comments"
## [13] "deployment.end.type"     "manipulation.type"      
## [15] "study.site"              "tag.beacon.frequency"   
## [17] "tag.manufacturer.name"   "tag.model"              
## [19] "tag.readout.method"

Note - that bizarro character ?.. in front of the first column is some strange Unicode encoding issue (which may or may not show up when you try to replicate this). We have to reload the data with an extra encoding option to remove it:

Note, also, that the original data file had columns that used dashes as separators, e.g.: tag-id, animal-id, animal-taxon, deploy-on-date, deploy-off-date, etc. These are transformed into periods when read into R, since the dash is a minus sign.

This data frame contains several many columns that aren’t in the inuvik@idData slot, compare:

names(inuvik@idData)
##  [1] "comments"             "death_comments"       "earliest_date_born"  
##  [4] "exact_date_of_birth"  "individual_id"        "latest_date_born"    
##  [7] "local_identifier"     "nick_name"            "ring_id"             
## [10] "sex"                  "taxon_canonical_name"

3.2 Merging with idData

We can merge all of the information in the reference data frame using the very powerful merge function. Note that you have to specify the columns by which you’re matching the data frames, in this case: local_identifier matches the tag.id. Note that we use all.x = TRUE, which retains only rows of the first data frame, but not necessarily all the rows of the second data frame, in order to to keep the same basic shape as the original @idData.

inuvik.idData.merged <- merge(inuvik@idData, inuvik.ref.data,
                              by.x = "local_identifier", 
                              by.y = "animal.id", 
                              all.x = TRUE)

This created a merged data frame, with, however, slightly different number of rows than @idData:

dim(inuvik.idData.merged)
## [1] 45 29
dim(inuvik@idData)
## [1] 43 11

This is no good! It happens because there are a few “animal.id”’s in the refrence data with multiple rows. Here’s some quick code to see which ones:

tail(sort(table(inuvik.ref.data$animal.id)))
## 
## BWC36186 BWC36187 BWC36188 BWC36189     BW26     BW34 
##        1        1        1        1        2        2

I’m not sure why there are multiple rows, but since they are associated with only a single row in the inuvik reference data, there might be some confusion! We need to eliminate the excess rows for the following to work.

which(inuvik.ref.data$animal.id == "BW26")
## [1] 10 29
which(inuvik.ref.data$animal.id == "BW34")
## [1] 16 30
inuvik.ref.data <- inuvik.ref.data[,-c(29,30)]
inuvik.idData.merged <- merge(inuvik@idData, inuvik.ref.data,
                              by.x = "local_identifier", 
                              by.y = "animal.id", 
                              all.x = TRUE)

Unfortunately, we can’t simply replace the @idData with this new data frame. When we do so, and convert to a data frame all the information is lost. See:

inuvik2 <- inuvik
inuvik2@idData <- inuvik.idData.merged
as.data.frame(inuvik2) %>% str
## 'data.frame':    69997 obs. of  50 variables:
##  $ sensor_type_id         : int  653 653 653 653 653 653 653 653 653 653 ...
##  $ comments               : logi  NA NA NA NA NA NA ...
##  $ habitat                : logi  NA NA NA NA NA NA ...
##  $ import_marked_outlier  : logi  NA NA NA NA NA NA ...
##  $ location_lat           : num  67.5 67.5 67.5 67.5 67.5 ...
##  $ location_long          : num  -132 -132 -132 -132 -132 ...
##  $ manually_marked_outlier: Factor w/ 2 levels "false","true": 1 1 1 1 1 1 1 1 1 1 ...
##  $ timestamp              : POSIXct, format: "2005-03-11 00:00:00" "2005-03-11 08:03:42" ...
##  $ update_ts              : Factor w/ 2 levels "2018-09-17 20:32:20.291",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ argos_calcul_freq      : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ argos_lc               : int  NA NA NA NA NA NA NA NA NA NA ...
##  $ deployment_id          : int  532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 ...
##  $ event_id               : num  7.41e+09 7.41e+09 7.41e+09 7.41e+09 7.41e+09 ...
##  $ tag_id                 : int  532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 ...
##  $ sensor_type            : Factor w/ 2 levels "Argos Doppler Shift",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ location_long.1        : num  -132 -132 -132 -132 -132 ...
##  $ location_lat.1         : num  67.5 67.5 67.5 67.5 67.5 ...
##  $ optional               : logi  TRUE TRUE TRUE TRUE TRUE TRUE ...
##  $ sensor                 : Factor w/ 2 levels "Argos Doppler Shift",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ timestamps             : POSIXct, format: "2005-03-11 00:00:00" "2005-03-11 08:03:42" ...
##  $ trackId                : Factor w/ 43 levels "BW26","BW27",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ local_identifier       : Factor w/ 43 levels "BW26","BW27",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ comments.1             : logi  NA NA NA NA NA NA ...
##  $ death_comments         : logi  NA NA NA NA NA NA ...
##  $ earliest_date_born     : logi  NA NA NA NA NA NA ...
##  $ exact_date_of_birth    : logi  NA NA NA NA NA NA ...
##  $ individual_id          : int  NA NA NA NA NA NA NA NA NA NA ...
##  $ latest_date_born       : logi  NA NA NA NA NA NA ...
##  $ nick_name              : Factor w/ 43 levels "IN0701","IN0704",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ ring_id                : logi  NA NA NA NA NA NA ...
##  $ sex                    : Factor w/ 1 level "f": NA NA NA NA NA NA NA NA NA NA ...
##  $ taxon_canonical_name   : Factor w/ 1 level "caribou": NA NA NA NA NA NA NA NA NA NA ...
##  $ ï..tag.id              : Factor w/ 41 levels "12179","12181",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ animal.taxon           : Factor w/ 1 level "Rangifer tarandus": NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.on.date         : Factor w/ 13 levels "2002-05-01 00:00:00.000",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.off.date        : Factor w/ 33 levels "2003-10-27 23:59:00.000",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ animal.sex             : Factor w/ 1 level "f": NA NA NA NA NA NA NA NA NA NA ...
##  $ animal.taxon.detail    : Factor w/ 1 level "caribou": NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.off.latitude    : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.off.longitude   : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.on.latitude     : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ deploy.on.longitude    : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ deployment.end.comments: Factor w/ 25 levels "can be stationary much before 1/10/2006, not quality data, mulfunction afer 5/8/2006",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ deployment.end.type    : Factor w/ 5 levels "dead","equipment-failure",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ manipulation.type      : Factor w/ 1 level "none": NA NA NA NA NA NA NA NA NA NA ...
##  $ study.site             : Factor w/ 1 level "Inuvik": NA NA NA NA NA NA NA NA NA NA ...
##  $ tag.beacon.frequency   : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ tag.manufacturer.name  : Factor w/ 1 level "Telonics": NA NA NA NA NA NA NA NA NA NA ...
##  $ tag.model              : Factor w/ 3 levels "ST-10","ST-14",..: NA NA NA NA NA NA NA NA NA NA ...
##  $ tag.readout.method     : Factor w/ 1 level "satellite": NA NA NA NA NA NA NA NA NA NA ...

All the new data is NA!?

3.3 Merging with movement data frame

So we have to “manually” merge the location data frame with the ORIGINAL reference data frame, using the trackID as the merging criterion. Thus:

inuvik.df <- as.data.frame(inuvik)
inuvik.df <- merge(inuvik.df, inuvik.ref.data, 
                   by.x = "trackId", by.y = "animal.id", all.x = TRUE)
str(inuvik.df)
## 'data.frame':    79987 obs. of  50 variables:
##  $ trackId                : Factor w/ 43 levels "BW26","BW27",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ sensor_type_id         : int  653 653 653 653 653 653 653 653 653 653 ...
##  $ comments               : logi  NA NA NA NA NA NA ...
##  $ habitat                : logi  NA NA NA NA NA NA ...
##  $ import_marked_outlier  : logi  NA NA NA NA NA NA ...
##  $ location_lat           : num  68.3 68.3 68.3 68.3 68.3 ...
##  $ location_long          : num  -133 -133 -133 -133 -133 ...
##  $ manually_marked_outlier: Factor w/ 2 levels "false","true": 1 1 1 1 1 1 1 1 1 1 ...
##  $ timestamp              : POSIXct, format: "2005-07-18 16:01:55" "2005-07-18 16:01:55" ...
##  $ update_ts              : Factor w/ 2 levels "2018-09-17 20:32:20.291",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ argos_calcul_freq      : num  NA NA NA NA NA NA NA NA NA NA ...
##  $ argos_lc               : int  NA NA NA NA NA NA NA NA NA NA ...
##  $ deployment_id          : int  532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 532845469 ...
##  $ event_id               : num  7.41e+09 7.41e+09 7.41e+09 7.41e+09 7.41e+09 ...
##  $ tag_id                 : int  532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 532618725 ...
##  $ sensor_type            : Factor w/ 2 levels "Argos Doppler Shift",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ location_long.1        : num  -133 -133 -133 -133 -133 ...
##  $ location_lat.1         : num  68.3 68.3 68.3 68.3 68.3 ...
##  $ optional               : logi  TRUE TRUE TRUE TRUE TRUE TRUE ...
##  $ sensor                 : Factor w/ 2 levels "Argos Doppler Shift",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ timestamps             : POSIXct, format: "2005-07-18 16:01:55" "2005-07-18 16:01:55" ...
##  $ comments.1             : logi  NA NA NA NA NA NA ...
##  $ death_comments         : logi  NA NA NA NA NA NA ...
##  $ earliest_date_born     : logi  NA NA NA NA NA NA ...
##  $ exact_date_of_birth    : logi  NA NA NA NA NA NA ...
##  $ individual_id          : int  532618724 532618724 532618724 532618724 532618724 532618724 532618724 532618724 532618724 532618724 ...
##  $ latest_date_born       : logi  NA NA NA NA NA NA ...
##  $ local_identifier       : Factor w/ 43 levels "BW26","BW27",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ nick_name              : Factor w/ 43 levels "IN0701","IN0704",..: 16 16 16 16 16 16 16 16 16 16 ...
##  $ ring_id                : logi  NA NA NA NA NA NA ...
##  $ sex                    : Factor w/ 1 level "f": 1 1 1 1 1 1 1 1 1 1 ...
##  $ taxon_canonical_name   : Factor w/ 1 level "caribou": 1 1 1 1 1 1 1 1 1 1 ...
##  $ ï..tag.id              : Factor w/ 41 levels "12179","12181",..: 26 2 26 2 26 2 26 2 26 2 ...
##  $ animal.taxon           : Factor w/ 1 level "Rangifer tarandus": 1 1 1 1 1 1 1 1 1 1 ...
##  $ deploy.on.date         : Factor w/ 13 levels "2002-05-01 00:00:00.000",..: 8 4 8 4 8 4 8 4 8 4 ...
##  $ deploy.off.date        : Factor w/ 33 levels "2003-10-27 23:59:00.000",..: 26 14 26 14 26 14 26 14 26 14 ...
##  $ animal.sex             : Factor w/ 1 level "f": 1 1 1 1 1 1 1 1 1 1 ...
##  $ animal.taxon.detail    : Factor w/ 1 level "caribou": 1 1 1 1 1 1 1 1 1 1 ...
##  $ deploy.off.latitude    : num  67.6 67.7 67.6 67.7 67.6 ...
##  $ deploy.off.longitude   : num  -132 -132 -132 -132 -132 ...
##  $ deploy.on.latitude     : num  67.7 67.6 67.7 67.6 67.7 ...
##  $ deploy.on.longitude    : num  -132 132 -132 132 -132 ...
##  $ deployment.end.comments: Factor w/ 25 levels "can be stationary much before 1/10/2006, not quality data, mulfunction afer 5/8/2006",..: 11 8 11 8 11 8 11 8 11 8 ...
##  $ deployment.end.type    : Factor w/ 5 levels "dead","equipment-failure",..: 1 4 1 4 1 4 1 4 1 4 ...
##  $ manipulation.type      : Factor w/ 1 level "none": 1 1 1 1 1 1 1 1 1 1 ...
##  $ study.site             : Factor w/ 1 level "Inuvik": 1 1 1 1 1 1 1 1 1 1 ...
##  $ tag.beacon.frequency   : num  150 151 150 151 150 ...
##  $ tag.manufacturer.name  : Factor w/ 1 level "Telonics": 1 1 1 1 1 1 1 1 1 1 ...
##  $ tag.model              : Factor w/ 3 levels "ST-10","ST-14",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ tag.readout.method     : Factor w/ 1 level "satellite": 1 1 1 1 1 1 1 1 1 1 ...

We now have ALL that ancillary data attached to the inuvik.df data frame.

Note, this is a BIG data frame! As a general R workflow recommendation, it is a good idea to save this object in R’s native format to quickly reload it as needed, e.g.:

save(inuvik.df, file = "./_data/InuvikProcessed.rda")

So that when you revisit the code you can quickly load the data into your namespace via:

load("./_data/InuvikProcessed.rda")

4 Plotting the data

Ok, enough messing with data. Let’s do some plotting! Note, we can load our processed data to skip a lot of the steps above:

load("./_data/InuvikProcessed.rda")

4.1 Basic plot

A useful tool for plotting structured data like this (x-y-id) is with ggplot. There are lots of references online, but here is an example:

require(ggplot2)
ggplot(inuvik.df, aes(location_long, location_lat, col = nick_name)) + geom_path() + geom_point()

4.2 GGmaps

There is a package called ggmap with allows us to use Google and other available map sources to improve our plots. There are several steps here:

load(file = "./_data/inuvikmap.rda")
  1. Load ggmap (or first install it if you don’t have it) and obtain a “basemap”, which requires a range and a zoom level.
require(ggmap)
latrange <- range(inuvik.df$location_lat) + c(-.5, .5)
longrange <- range(inuvik.df$location_long) + c(-.5, .5)
## inuvik.map <- get_map(location = c(longrange[1], latrange[1], longrange[2], latrange[2]), maptype = "terrain", source = "google")

There are lots of options for the maptype, e.g. “terrain”, “satellite”, “roadmap”, “hybrid”. The choice depends on the information you want to communicate (and how much you want to foreground your tracks).

Note: if your data is a Move object, you can use the bbox() command to automatically obtain the bounding box of the data range code as follows:

require(move)
inuvik.map <- get_map(bbox(inuvik), source="google", maptype="satellite")
  1. There will be no output - you just created an objet called inuvik.map, which you can view with the ggmap command:
ggmap(inuvik.map)

then add your data onto that map

ggmap(inuvik.map) + 
    geom_point(data = inuvik.df, mapping = aes(x = location_long, y = location_lat, col = nick_name), alpha = 0.5, size = 0.5) + 
    geom_path(data = inuvik.df, mapping = aes(x = location_long, y = location_lat, col = nick_name), alpha = 0.5, size = 0.5) +
    labs(x="Longitude", y="Latitude", main = "Inuvik Caribou")

BAD NEWS (UPDATE OCTOBER 27, 2018): The code above is not guaranteed to work since Google has decided to make their maps API no longer (entirely) free. You can still do this, but you will need an “api_key”, for which you need to enable billing (see here: https://developers.google.com/maps/documentation/javascript/get-api-key). This is a major drag. However, there are other cool options …..

4.3 Mapview

There is a very powerful open-source scripting library called leaflet, which allows for the creation of interactive maps with a large range of backgrounds. It’s remarkable easy to use from R, but there is some prep work that needs to be done.

The key packages:

require(leaflet)
require(mapview)

4.3.1 First, simple features

We are going to convert our inuvik data into a special kind of object (to which we will be returning to often) called a simple feature, which is basically a data frame where one column contains all of the spatial information. It is easy to create a simple feature object out of a move object. Below we do the conversion, and merge it with some extra useful information from the inuvik.df:

require(sf)
inuvik.sf <- st_as_sf(inuvik) 
inuvik.sf <- merge(inuvik.sf, inuvik.df[,c("timestamp","deployment_id","event_id","nick_name")])

Simple features allow us to do some nicely projected mapping. For example:

ggplot(subset(inuvik.sf, nick_name == nick_name[1]), aes(location_long, location_lat, col = nick_name)) + geom_sf() + geom_path()

But if you wanted it projected to, e.g., Canada Albers Equal Area Conic (ESRI:102002)

ggplot(subset(inuvik.sf, nick_name == nick_name[1]) %>% st_transform(102002), aes(col = nick_name)) + geom_sf()

To turn the above into a path, which is what we really want, we need to “collapse” the POINT geometry of this caribou, into a single LINES geometry. That’s done as follows (and I confess I don’t entirely understand the sequence of dplyr commands, but it doesn’t seem to work without the grouping AND the summarizing):

require(dplyr)
inuvik.line <-inuvik.sf[,"nick_name"] %>% group_by(nick_name) %>% summarize(do_union=FALSE) %>% st_cast("LINESTRING")
ggplot(inuvik.line %>% st_transform(102002), aes(col = nick_name)) + geom_sf() + ggtitle("Inuvik barrenground caribou: Canada Lambert Conformal Conic")

4.3.2 Interactive mapping

Now that the LINES simple feature data frame is set up, plotting this is as easy as:

mapview(inuvik.line, zcol = "nick_name", map.types = "Esri.WorldPhysical")

Note: this is an interactive map. You can zoom it, drag it, and scroll over the paths a nd see the ID of the animal. “Esri.WorldPhysical” is just one of many possible maps. Here’s a link to a bunch of other ones. Here’s another, more stylized one:

mapview(inuvik.line, zcol = "nick_name", map.types = "Esri.NatGeoWorldMap")

Easy! And quite useful, I think.

5 Complete code

Click below to show complete code from this document.

knitr::opts_chunk$set(echo = TRUE, cache = TRUE, warning = FALSE, message = FALSE)
pcks <- c("magrittr", "rmarkdown")
a <- lapply(pcks, require, character = TRUE)
require(move)
mylogin <- movebankLogin(username="Elie Gurarie", password="Caribou")
## mylogin <- movebankLogin(username="MyUserName", password="MyPassword")
inuvik <- getMovebankData("ABoVE: NWT Inuvik Boreal Woodland Caribou", login = mylogin)
## save(inuvik, file = "./_data/inuvik.rda")
load(file = "./_data/inuvik.rda")
inuvik
slotNames(inuvik)
## head(inuvik@data)
head(inuvik@data) %>% kable
inuvik@proj4string
inuvik@license
inuvik@citation
## inuvik@idData
require(knitr)
kable(inuvik@idData)
table(inuvik@trackId)
inuvik.df <- as.data.frame(inuvik)
names(inuvik.df)
inuvik.df[,c("trackId", "local_identifier", "nick_name", "deployment_id", "tag_id","individual_id")] %>% unique
## setwd("C:/Users/Guru/box sync/caribou/NWT/rsf/primers")
## load("./_data/inuvik.rda")
inuvik.ref.data <- read.csv("./_data/ABoVE_ NWT Inuvik Boreal Woodland Caribou-reference-data.csv")
names(inuvik.ref.data)
## inuvik.ref.data <- read.csv("./_data/ABoVE_ NWT Inuvik Boreal Woodland Caribou-reference-data.csv", fileEncoding="UTF-8-BOM")
## names(inuvik.ref.id)
names(inuvik@idData)
inuvik.idData.merged <- merge(inuvik@idData, inuvik.ref.data,
                              by.x = "local_identifier", 
                              by.y = "animal.id", 
                              all.x = TRUE)
dim(inuvik.idData.merged)
dim(inuvik@idData)
tail(sort(table(inuvik.ref.data$animal.id)))
which(inuvik.ref.data$animal.id == "BW26")
which(inuvik.ref.data$animal.id == "BW34")
inuvik.ref.data <- inuvik.ref.data[,-c(29,30)]
inuvik.idData.merged <- merge(inuvik@idData, inuvik.ref.data,
                              by.x = "local_identifier", 
                              by.y = "animal.id", 
                              all.x = TRUE)
inuvik2 <- inuvik
inuvik2@idData <- inuvik.idData.merged
as.data.frame(inuvik2) %>% str
inuvik.df <- as.data.frame(inuvik)
inuvik.df <- merge(inuvik.df, inuvik.ref.data, 
                   by.x = "trackId", by.y = "animal.id", all.x = TRUE)
str(inuvik.df)
## save(inuvik.df, file = "./_data/InuvikProcessed.rda")
## load("./_data/InuvikProcessed.rda")
load("./_data/InuvikProcessed.rda")
require(ggplot2)
ggplot(inuvik.df, aes(location_long, location_lat, col = nick_name)) + geom_path() + geom_point()
load(file = "./_data/inuvikmap.rda")
require(ggmap)
latrange <- range(inuvik.df$location_lat) + c(-.5, .5)
longrange <- range(inuvik.df$location_long) + c(-.5, .5)
inuvik.map <- get_map(location = c(longrange[1], latrange[1], longrange[2], latrange[2]), maptype = "terrain", source = "google")
## save(inuvik.map, file = "./_data/inuvikmap.rda")
## require(move)
## inuvik.map <- get_map(bbox(inuvik), source="google", maptype="satellite")
ggmap(inuvik.map)
ggmap(inuvik.map) + 
    geom_point(data = inuvik.df, mapping = aes(x = location_long, y = location_lat, col = nick_name), alpha = 0.5, size = 0.5) + 
    geom_path(data = inuvik.df, mapping = aes(x = location_long, y = location_lat, col = nick_name), alpha = 0.5, size = 0.5) +
    labs(x="Longitude", y="Latitude", main = "Inuvik Caribou")
require(leaflet)
require(mapview) 
require(sf)
inuvik.sf <- st_as_sf(inuvik) 
inuvik.sf <- merge(inuvik.sf, inuvik.df[,c("timestamp","deployment_id","event_id","nick_name")])
ggplot(subset(inuvik.sf, nick_name == nick_name[1]), aes(location_long, location_lat, col = nick_name)) + geom_sf() + geom_path()
ggplot(subset(inuvik.sf, nick_name == nick_name[1]) %>% st_transform(102002), aes(col = nick_name)) + geom_sf()
require(dplyr)
inuvik.line <-inuvik.sf[,"nick_name"] %>% group_by(nick_name) %>% summarize(do_union=FALSE) %>% st_cast("LINESTRING")
ggplot(inuvik.line %>% st_transform(102002), aes(col = nick_name)) + geom_sf() + ggtitle("Inuvik barrenground caribou: Canada Lambert Conformal Conic")
require(mapview)
mapview(inuvik.line, zcol = "nick_name", map.types = "Esri.WorldPhysical")
require(mapview)
mapview(inuvik.line, zcol = "nick_name", map.types = "Esri.NatGeoWorldMap")