Difference between revisions of "File Types"

From Anatrack Ranges User Guide
Jump to: navigation, search
Line 44: Line 44:
  
 
These must be set to convert UTM coordinates to latitude-longitude.
 
These must be set to convert UTM coordinates to latitude-longitude.
 
  
 
==== Range attribute variables ====
 
==== Range attribute variables ====

Revision as of 11:45, 10 November 2014

Introduction

All data files in Ranges contain text, and can therefore be viewed in text editors, except for raster maps (.rst) and image files (.ima) which are byte arrays. Locations files (.loc), vector files (.vep, .vel, & .ves , for point, line and shape maps), edge files (.edg, of range outlines), utilisation files (.uti), incremental files (.inc) and survival files (.srv) are mainly to be used as inputs for other Ranges analyses. They can also be viewed in a spreadsheet, but the data are arranged to save space rather than for ease of interpretation. Outputs from modelling analyses, RADA files (.rda) and Kaplan Meier Graph files (.kms) have no map data and are only useful for displaying output data and plots within Ranges. Ranges also outputs files for use in other software, e.g. ‘.csv’ files that can be used spreadsheets (simply double click on them to open in Excel), and ESRI shapefiles that can be opened in ArcView or ArcGIS.

Location Files

Location files (.loc) are the main starting point for analyses. They contain the point location data and associated attribute information. Locations are stored in a series of ranges, which may represent an individual, a social group or a particular time period. Location files can be created from scratch, or imported. To ensure that Ranges is suitable for tracking all animals (from ants to elephants), each file must include information about the tracking resolution, scale and whether a focal site (e.g. nest, den etc.) should be included in the range.

Location coordinates

The locations themselves must be recorded in a flat projection, such as UTM ( or for Britain the OS National Grid ). Ranges can convert to UTM from latitude-longitude during the [[Input & Graphics#Import|import procedure and an Excel spreadsheet for doing this is available cheaply from [[1]]. GPS receivers usually have the option to output coordinates in UTM rather than Lat/Long.

Location coordinates can be decimals, but they cannot be negative. If you have locations with negative coordinates you will need to first transform your data by adding a number to the Eastings and or Northings such that all of the coordinates become positive. (The added number will have to be greater than or equal to the magnitude of the most negative easting or northing). This can be accomplished in a spreadsheet such as Excel.

File attribute variables

Tracking Resolution

The resolution depends on the accuracy of your tracking. Accuracy is sometimes cited as the standard deviation of locations obtained in a trail. Greater comparability with other statistical selection criteria would be to use the 95% confidence distance. For tracking with a 2-3 element Yagi, a 1-in-10 rule is acceptable: if you make triangulations at up to 10m from a small animal, the tracking resolution is 1m, or 10m at 100m. For Yagi antennas with more elements, a 1-in-20 rule can apply, i.e. a resolution of 100m for triangulations at up to 2km from an animal. The resolution with GPS may vary from 10-50m depending on terrain and other conditions. For ARGOS satellite tracking it may be closer to 1 km and depend on location class. Tracking resolution is used to:

  1. Set the width of the boundary strip that is included in polygon edges and areas. The boundary strip is half the value of the resolution e.g. 0.5m if the resolution is 1m. This means that a coordinate with a real position between 10.5 and 11.5, can be entered as 11. If you wish to suppress the boundary strip you can set the tracking resolution to 0. This is useful if you need analysis results comparable with a system lacking boundary strip estimation.
  2. Estimate the size of a grid cell plotted around single outlying locations in concave polygon analyses, and multiple locations at one site in cluster analyses. The width of the grid cell is equal to the resolution.
  3. To define the truncation distance in one type of harmonic mean contouring. In this case, the allocation of an inappropriate resolution can substantially affect range areas and statistics.

Scale

Scale is the number of metres represented by each coordinate unit. A scale of 10m means that each coordinate unit (e.g. 8,8) is 10m from the next (e.g. 8,9). Scaling greater than 1 allows you to type in data with fewer digits. For example, you can use a scale of 100m and type 213 instead of 21300.

Warning! The scale parameter must match the scale parameter in any vector or raster maps with which you wish to compare your data.

The scale parameter does not influence the map display in Input & Graphics (which uses the untransformed coordinate values), so a location file may appear to overlay a vector map file, but if they have different scale parameters they may not overlay in, for example, habitat analysis.

If a file is exported to an ArcView shapefile the coordinates are multiplied by scale to convert them to metres. If a file is exported to a text file the coordinates are not multiplied by the scale parameter and the scale parameter will have to be re-entered if the text file is re-imported into Ranges.

Include Nest/Focal Site In Locations

If you choose yes, the focal site for each range will be included as a location in later analyses.

Coordinate System Datum Ellipsoid

If your data is in UTM (Universal Transverse Mercator) coordinates, it can be converted to latitude-longitude coordinates for display on Google Maps or for export to KML. The Coordinate System Datum Ellipsoid, the ellipsoid used to model the shape of the Earth, must be set to achieve this. Most UTM systems use WGS84.

UTM Latitude Zone and UTM Longitude Zone

These must be set to convert UTM coordinates to latitude-longitude.

Range attribute variables

A file may contain thousands of ranges from different individuals, or from the same individual during different tracking sessions, and these must be easily identified. Saving multiple ranges in the same file can make analyses much easier as you only need to run once per file. It is easy to select the ranges you wish to look at during the analysis (e.g. Only the males). You can define how you want to identify the ranges, but the defaults are ID number, Age, Sex, Month and Year. These are explained below:

All seven Range attribute variables are stored as numeric values (for sorting purposes). The first five variables (ID, Age, Sex, Month and Year) must be integers, whilst the focal site coordinates can be decimals. The Age and Sex variables can be given one letter labels to identify animals on the screen.

The default Range attribute variable labels are as follows:

Age 1=J, 2=Y, 3=A (for Juvenile, Yearling, Adult)
Sex 1=M, 2=F (for Male, Female)

There are default values for each range variable, newly created ranges will start with these values.

ID sequence number, which increments by one for each consecutive range
AGE  ? (coded as 0)
SEX  ? (coded as 0)
MONTH (missing, coded as -9)
YEAR (missing, coded as -9)
FOCAL SITE COORD E (missing, coded as -9)
FOCAL SITE COORD N (missing, coded as -9)


A typical label might be displayed as AF21 6/01 for Adult Female 21 tracked from (or during) June 2001. Missing variables appear on results screens as ?. If you only define IDs, the same label would be ??21 ?/?.

ID should be unique for each range ( e.g. you shouldn’t have a male and female both with ID 21). By using import and export and changing the ID you can subdivide your data in different ways, e.g. to look at range changes over time.

Location qualifying variables (LQVs)

It is often helpful to know what an animal is doing or what habitat an animal is in at each location, or when the location was recorded, or even a score of how accurate the location was. This is achieved by using Location Qualifying Variables, of which there can be up to 50. It is easy to select the locations you wish to look at during the analysis (e.g. excluding locations when the animal is sleeping).

Time LQVs

Labels for times must be capital letters, in combinations of 2, 4 or 6 letters of the following : YY, MO, DD, HH, MI, SS. e.g. YYDD, HHMI, YYMODD, DDMOYY. There is a limit of 6 characters in the time LQVs so that 010417 is OK for April 17 2001, but 20010417 is not. A useful tip. If time variables are entered as "combinations" in descending order, (eg. YYMODD or HHMI), you can use these in Make Selections to select runs of locations that cross period boundaries. Using single labels, it is difficult to select locations between 15 December 1994 and 1 February 1995 while excluding locations between 1-14 December 1994 and those in January 1994. However, combination labels make it easy to select locations between 941215 and 950201.

Accuracy ellipse LQVs

LQVs can also be used to represent accuracy ellipses in which case the labels must be defined by capitals as MARAD, MIRAD, THETA or X-VAR, Y-VAR, COVAR, which represent either the maximum radius, minimum radius and inclination of an ellipse or the variance and covariance of its distribution (see the output statistics from location analysis, ellipses for more details).

Labels for other LQVs can be upper or lower case and should ideally have 5 letters (e.g. ACTIV) for them to show best in displays. Please note that although LQV labels and their range of values are listed to help make selections during analyses, you still need to record (outside of Ranges) the numerical codes of these 5-letter labels (eg. ACTIV, 1=resting, 2=feeding, 3=preening).

Vector Files

Vector files (.vep, .vel, .ves) can contain either point, line or shape information and are used for mapping habitats. Coordinates are stored in groups, each group representing a single shape, line or point group. Each group is assigned to a category which has an associated label and colour. Groups may have the same No. and ID, this is useful for the representation of holes and secondary polygons within shapes. A scale parameter defines the m per coord unit.

Vector file attributes can be set in the vector properties window, which is displayed when new...vector or ‘modify’ is pressed from the main panel. Vector files can be imported as text files with columns containing point coordinates or from ArcView shapefiles.

Vector files can be used in the following ways in Ranges:

  1. Background maps in location analyses: Point, line or shape files can be used as backgrounds for location analyses in which case they are plotted in grey. Grey is used for maps in this option to avoid confusion with the several colouring options for range edges.
  2. In Habitat analyses: Vector shapes and points are also used in colour in habitat analyses.
  3. In Interaction analyses: Shapes can be used to define sampling areas in interaction analyses. Points can be used to see, for example, how close a bird goes to nest sites or feeding areas, allowing objective assessments of association or avoidance.
  4. In Location Analyses: Vector line files formatted as midline files can be used in midline analyses to define potential routes between locations (e.g. for fish in rivers).

On-screen digitising to create Vector files

Use new to create an empty vector file.Points can be digitised by holding down the CTRL key and clicking the left mouse button, perhaps following an image background.

Vector file type - points, lines or shapes

Points files have no connections between coordinates. In lines files, lines are drawn between the coordinates within each line. In shapes files, coordinates in the same shape are connected by lines and the resultant shape is filled with colour.

Coordinates for shapes must be entered in a clockwise sequence, without lines that cross and must finish by repeating the first set of coordinates.

Holes can be added to shapes: first select the boundary shape and click hole. The ID and category of a hole is the same as its shell shape and cannot be edited.

Vector shapefiles including holes can be digitised and the coordinates copied to the following shape to fill in the hole.

Scale of coordinate units

The scale specifies the number of metres that each coordinate unit represents. For example if the scale is 1, a location at 1,1 will be 5m from a location at 1,5, with a scale of 10 this distance would be 50m.

Coordinate System Datum Ellipsoid

If your data is in UTM (Universal Transverse Mercator) coordinates, it can be converted to latitude-longitude coordinates for display on Google Maps or for export to KML. The Coordinate System Datum Ellipsoid, the ellipsoid used to model the shape of the Earth, must be set to achieve this. Most UTM systems use WGS84.

UTM Latitude Zone and UTM Longitude Zone

These must be set to convert UTM coordinates to latitude-longitude.

Vector categories

The add category button can be used to add new categories. Within the table the category label can be edited by double clicking in its cell. Category labels can be words or numbers. If the labels are numbers (e.g. density values), they can be decimals of any length. If you use numbers, for example to denote density of food in particular areas, Habitat content of ranges will produce an average value from the proportion of the different areas in each range.

Colour can be edited by a single click within the cell that will bring up a colour chooser. You will be forced to choose another colour if the colour chosen is too close to that of an existing category.

There is a limit of 50 vector categories within Ranges.

For ArcView shapefiles that have > 50 categories, in ArcView you can create two or more maps with a subset of the categories with the following procedure:

  1. Select the theme containing the whole map
  2. Theme, Query, category < 50th category
  3. Theme, Start editing
  4. Edit, Delete features
  5. Theme, Save edits as, new filename
  6. Repeat the above, changing the query in step 2) to >= 50th category

The maps can then be imported separately into Ranges.

Raster Files

Raster files (.rst) are files that store information, usually habitat, as a grid of cells, each containing a single map value. They are generally used for land cover information, particularly when it is derived from remotely sensed data from satellites or aerial photography. The size of each raster cell is defined and cannot be altered, which means that cells are displayed and analysed as large squares relative to fine-scaled range polygons, whereas vector maps retain angular shapes at any scale. However, analyses that use raster maps are faster than shapes over large areas. Moreover, makers of mapping systems tend not to export vectors in a convenient format for other packages, but will export raster files. Be careful to choose a raster cell size that provides adequate detail for analysis but remember that large raster maps are memory-hungry.

Ranges stores raster data as a byte array with a text appendix (these files are not text files and so cannot be viewed using a text editor). Although Ranges currently only handles 15 raster categories during analysis, maps with many more categories can be used. Categories can be easily combined, e.g. Orchard could be combined with deciduous woodland, and most analyses end up being based on less than 10 combinations. It is also simple to set categories to be unclassified, so that one set of 15 types are used for a run, and another set for a repeat run.

Raster files can be created from scratch or imported from gridascii files. The byte format in Ranges is typically about a third the size of the equivalent gridascii file, and therefore requires less memory.

Older versions of Ranges used to limit the area of a raster file you could view at a time. With increased computer memory this is no longer necessary however if you open a raster file with more than 10 million cells, the cell values table will not be displayed.

If you have problems loading large rasters, it is likely that you are running Ranges in a Java Runtime Environment (JRE) with limited memory. It is straight forward to increase the memory reserved for Ranges.

Further instructions on setting up raster files can be found here: Raster File Setup.

Edge Files

Edge files (.edg) store the coordinates of home-range shapes generated in location analyses.

They can store multiple ranges (which may represent different individuals or different tracking sessions), and for each range can store multiple cores representing different levels of use.

Edge files can be viewed and exported from the main window, but edge shapes cannot be edited there. When opened in Input & Graphics (as either the primary file or a background map), if you hold the mouse pointer over the filename a description of the analysis used to create the analysis will be displayed.

Edge files are the principal input for Overlap analyses.

Utilisation Files

Utilisation files (.uti) are created in location analysis and contain the areas of range cores at 5% intervals (e.g. for convex polygons). Utilisation plots can be viewed and printed from Input & Graphics, by selecting the open...Utilisation files option.

Incremental Files

Incremental files (.inc) are created in location analysis and contain the areas of ranges created from sub-samples of locations, starting from the first three and finishing with the entire set (e.g. for clusters). Incremental plots can be viewed and printed from Input & Graphics, by selecting the open...Incremental file option.

Statistics Files

Many analyses have an option within Run Specifications for outputting a statistics file. They are in CSV (comma-separated variable) format and are stored with the .csv extension. They have column headers and can be opened in the Ranges Statistics window, double-clicked to open in Microsoft Excel or imported to an alternative spreadsheet.

Esri Shapefiles

Esri shapefiles can be exported from Input & Graphics, and can be opened in ArcView, ArcGIS and other GIS packages. The ESRI shapefiles produced each consist of a minimum three files with the same root, .shp, .shx and .dbf, all three files are needed for it to be opened in another package. Shapefiles can also be imported into Ranges from Input & Graphics.

Image Files

Image files (.ima) can be used as backgrounds. They are created from standard image files, JPEG, PNG, GIF or bitmap, by importing into Ranges and aligning with a coordinate system.

RADA Files

RADA files (.rda) are the output of RADA analyses and contain habitat core data for the RADA plots. They cannot be edited.

Survival Files

Survival files (.srv) contain animal survival data and are used as the input to Kaplan Meier Survival analyses. Survival files have no location data, and therefore no map, but a number of extra range attribute variables in order to make survival analysis possible. These are Start Day, Start Month and Start Year, End Day, End Month and End Year, and Fate Code, the fate of the animal on the end date, which can be ? (Unknown), Lost, Lived or Died.

Kaplan Meier Survival Graph Files

Kaplan Meier Survival Graph files (.kms) are the output of Kaplan Meier analyses and contain data for the Kaplan Meier plots. They cannot be edited.