Tools to visualize movement data (e.g. from GPS tracking) and temporal changes of environmental data (e.g. from remote sensing) by creating video animations.
This is an R package providing tools to visualize movement data by creating path animations from geo-location point data. The package is under ongoing development. The moveVis package is working hand in hand with the move package by using the move and moveStack class and the raster package. It is based on a ggplot2 plotting architecture and relys on the libraries ImageMagick, ffmpeg and libav. To be informed about updates, new features and the current version, visit news.movevis.org.
This is the official moveVis R package repository, including beta code versions before submitted to CRAN. For operational use of moveVis, please use the current stable CRAN version of moveVis.
To install stable version from CRAN, please execute:
install.packages('moveVis')To install the development version from this GitHub repository, please execute:
devtools::install_github("16EAGLE/moveVis")You can use moveVis with any move or moveStack object. This guide shortly explains how to prepare your own geo-location point data for the animate_move() function by creating a move class object from a data.frame class object. As an example, the provided example data (data.frame) are used. Instead, you could use any similar prepared data of yours.
moveVis requires at least one of the three external libraries 'ffmpeg', 'libav' and/or 'ImageMagick'. They support different types of output formats (gif, mov, mp4 etc.). If you have 'ImageMagick' and either 'ffmepg' or 'libav' installed, you can use all output formats supported by moveVis.
Run get_libraries() to find out, which libraries are installed on your system:
get_libraries()On many Linux distributions, ImageMagick and FFmpeg are preinstalled or can be installed via the package manager. On Ubuntu, use sudo apt-get install imagemagick to install ImageMagick containing the convert tool and sudo apt-get install ffmpeg to install FFmpeg containing the equally called ffmpeg tool.
You can download and install both ImageMagick and FFmpeg from https://www.imagemagick.org/script/download.php as binary for Windows. Make sure that you select "Install FFmpeg" and "Install legacy utilities (e.g. convert)" during the installation process. get_libraries will recognize both libraries only, if their commands can be called from the command line. Make sure that you select the related option during the installation process.
Visit https://www.macports.org/install.php and follow the instructions there to download and install MacPorts. Execute sudo port install ImageMagick to install the latest binary release of ImageMagick for Mac. Visit https://www.ffmpeg.org/download.html to download and install the latest binary release of FFmpeg for Mac.
After the required libraries are once installed, restart your R session. First, load moveVis and the move package:
#Load packageslibrary(moveVis)library(move)and then run get_libraries() to check, if the installed tools are recognized by moveVis. If so, everything is set for starting to create your first moveVis animation. If the installed tools are not recognized automatically, you can provide the paths to the tools manually (see argument conv_dir in the manual of animate_move())
You will need to load the example data for this tutorial:
#Load data (data.frame) (or use your own as data.frame)data("move_data")As the provided example data, your data.frame needs to have at least three columns: two columns for your coordinates (here "lat", "lon") and one for the date/time stamp (here "dt"). The date/time stamps need to be converted to POSIXct as follows:
move_data$dt <- as.POSIXct(strptime(move_data$dt, "%Y-%m-%d %H:%M:%S", tz = "UTC"))Your movement data need to be provided as move class objects to the animate_move() function. For each individual movement path you want to display simultaniously within a single animation, you will need one move class object. The move class objects per path are provided as a list. If your data.frame contains several individuals (e. g. differentiable by a "individuals" column, as the example data.frame does), then subset the data per individual and store the namings. If you just want to display a single path, you do not have to do this.
#Create new move class object list by individualdata_ani <- split(move(move_data$lon, move_data$lat, proj=CRS("+proj=longlat +ellps=WGS84"), time = move_data$dt, animal=move_data$individual, data=move_data))get_libraries() returns the library commands that are needed by the animate functions. Just save them to a variable that you can later pass to the animate function so that it knows how to call the extern library commands. You can also call get_formats() to see all output formats, you can choose from.
#Get libraries conv_dir <- get_libraries() #Find out, which output file formats can be usedget_formats() Last, you need to specify the output directory path and you can specify some optional variables such as the animation title (for details on all the arguments of animate_move() , read the animate_move() help).
#Specify output directoryout_dir <- paste0(getwd(),"/test") #Specify some optional appearance variablesimg_title <- "Movement of the white stork population at Lake Constance, Germany"img_sub <- paste0("including individuals ",indi_names)img_caption <- "Projection: Geographical, WGS84; Sources: Movebank 2013; Google Maps"Finally, you are now prepared to call animate_move(), which will have to work for a while depending on your input. Here, for demonstrational purposes, we use frames_nmax set to 50 to force the function to only produce 50 frames and then finish the animation, regardless how many input points you provided. Set log_level to 1 to be informed of anything the function is doing. Set out_format to "mov" to get a .mov video output file.
#Call animate_move()animate_move(data_ani, out_dir, conv_dir = conv_dir, tail_elements = 10, paths_mode = "true_data", frames_nmax = 50, img_caption = img_caption, img_title = img_title, img_sub = img_sub, log_level = 1, out_format = "mov")After the function is finished, check the output directory. Retry everything with different settings and modes, described in the function manuals. Further examples and explanations are provided within the function manuals.
moveVis is being developed and maintained by Jakob Schwalb-Willmann. For bug reports, please use https://github.com/16eagle/movevis/issues to contact me. Feature requests and other contributions are also welcome.
The Department of Remote Sensing of the University of Würzburg has developed other R packages that might interest you:
For other news on the work at at the Department of Remote Sensing of the University of Würzburg, click here.
This initiative is part of the Opt4Environment project and was funded by the German Aerospace Center (DLR) on behalf of the Federal Ministry for Economic Affairs and Energy (BMWi) with the research grant 50 EE 1403.
Future implementations (not yet implemented!)
Reorganizing standard basemap usage by moveVis
Updating unit tests for CRAN checks
Windows library detection bugs fixed (dev. version)
video support, automatic time harmonization, bug fixes (CRAN version)
adding frames_layout, static_data etc., improvements to workflow
adding animate_stats() and stats arguments for animate_move()
fixing major bug
adding frame_width and frame_height arguments
fixing important bug; adding animate_raster()
initial release
This document should provide a broad overview on changes that are applied to the moveVis R package. There is no warranty for completeness, since minor changes might not be included. All improvement and feature descriptions are bundled per release version. The document is currently maintained by Jakob Schwalb-Willmann.