NMMU logo
Movement Ecology Workshop 2015

Part A: Data Preparation

Back to the roadmap home      Next to Part B: Data Analysis I

The first few steps apply to both raw location data and raw ancilliary data (e.g., from TDRs, Accelerometers, Videos, HRMs, etc) in parallel, before finally integrating them.



In the field.

If you are using multiple devices, have a "sync event", such as dunking all devices simultaneously into a bucket of icewater at a specific time. Do this both prior to deployment and post retrieval.



In your file manager and MS Excel, etc.

  • Ensure the file names are consistent (e.g., Prefix_SSSS_DDDDDDDD_PPPP_NNNN.ext where Ss are the species name, Ds are the date format as YYYYMMDD, Ps are the tagging place and Ns are serial numbers)
  • Ensure the file format is consistent (e.g., data columns, column formats - particularly for date and time)
  • Construct a metadata file



In QGIS, ArcGIS, IgorPro, R and/or Excel.

  • Start with basic plots of the data
    • Spatial maps
    • Time series plots of latitude and longitude
  • Remove obvious outliers
  • Remove duplicate points
  • Remove land/sea points for marine/terrestrial animals
    (this will be data dependent and/or species dependent and/or question dependent; if you are using the MKDE/BRB method, use 3*hmin/3*hmax as a guide for how far landward/seaward of the boundary you need to clean up)
  • Remove pre-deployment and post-retrieval data
  • Project the data in R or one of the GIS programmes
  • Convert the date and time to POSIXct format in R using the trip package or adehabitatLT
  • Create a trajectory (i.e., convert to ltraj format) and call the summary

Note:  Be careful to record what you have done, and don't permanently modify the raw data - save a copy of it somewhere (consider repositories like GitHub)
Note:  Be very careful with file versions

Recommended references: Zuur et al. 2010 A protocol for data exploration to avoid common statistical problems



  • Remove improbable points using:
    • speed filters
    • distance filters
    • kalman filters
    • mixture models
  • Tools and options in R include the argosfilter package and the mcconell filter in the trip package.
  • Those with Argos data can make use of the Douglas-Argos filter as implemented in Movebank (see the YouTube movie for a demonstration/tutorial)

Note: You may have to fill in gaps before integration (see below). Be aware of the difference between interpolating and rediscretizing.



In R, MS Excel, etc.

  • Work with a standardize time (e.g., GMT)
  • Interpolate positions (depending on the analysis)
  • Match location data and ancilliary data

Back to the roadmap home      Next to Part B: Data Analysis I


Nelson Mandela Metropolitan University

Telephone: 041-504 1111 | Fax: 041-504 2574 / 2731

E-mail: info@nmmu.ac.za

PO Box 77000 | Nelson Mandela Metropolitan University

Contact usPort Elizabeth | 6031 | South Africa

YouTube Twitter Facebook Connect@nmmu


© NMMU 2005-2014

A-Z Index

NMMU site map