This Chapter contains information about how to in- and export data from PCRaster. It discusses how to convert data from other GIS to one of the kinds of data used in PCRaster (Creation of a PCRaster map; data import) and how to export data from PCRaster (Data export from a PCRaster map). Other database management operations are also described, such as how to cut or join together maps and how to change location attributes or the data type of maps (Cutting or joining maps; changing geographical location attributes or data types). Furthermore it will be explained how to attach a legend to a map (Attaching a legend to a PCRaster map) and how to display PCRaster maps on the computer screen and print a map (Screen display, hard copy output of PCRaster maps).
The other sorts of data used in PCRaster (tables, time series and point data column files) are ascii formatted. Unlike PCRaster maps, these can easily be created, analyzed and edited with other software packages. Tables: database management and Time series: database management explain the creation of tables and time series, respectively. Creation and conversion of point data column files is described in Point data column files: database management. The above mentioned sections do not describe the formats of these sort of data. The format of PCRaster maps (PCRaster maps), tables (Table format), time series (Time series format) and point data column files (Point data column file format) are described succesively.
Global options gives a list of global options used in PCRaster. Global options are set only once (for instance at the beginning of a project) and will affect all operations to which they are relevant. Local options which are used each time a operation is done, must always be redefined.
If you start a project and want to use PCRaster maps for analyzing your data, first make an empty clone map to define the geographical and cartographical location attributes and set this map as global clone map with the global option –clone. To make the clone map, use the operator mapattr; the location attributes can be entered using the menu given by mapattr. This map can be used as a clone (mask map) for all data sets you want to import. A data type and its cell representation can also be attached to the clone map using mapattr. This data type and cell representation will by default be assigned to the data that are imported, but the type of the imported data can also be specified during the import.
The choice of geographical location attributes must be based upon the spatial characteristics of the data set you want to import to the PCRaster map format and upon whether x,y coordinates are attached to the data in the data set or not. There are two possibilities for data import:
The data file you want to import contains x and y coordinates with data values. In this case the location attributes of the PCRaster map must be chosen in correspondence with the spatial distribution of the data given by the x and y coordinates. If the data are regularly spaced on a rectangular grid, you probably want location attributes that match the set of x and y coordinates of the data set. If the data are irregularly spaced we advise to choose a map size of the smallest bounding rectangle (or somewhat larger) that comprises the study area, as shown in Location attributes used to define the spatial characteristics of a PCRaster map.. In this Figure a rectangle has been chosen that is rotated with respect to the real coordinate system (positive angle). Quite often, it is better not to rotate the map, and to choose an unrotated (smallest) rectangle. Rotation may result in a map that better fits the shape of the study area. But remember that rotation has an important effect which may be very clumsy: rotation will not only result in a rotated map but of course also in a grid of cells that is rotated with respect to the real coordinate system. As a result, cells that are in one row on the PCRaster map will not have the same y coordinates; the same holds for cells in one column and their x coordinates.
Two sorts of column files with x and y coordinates may be imported: a column file in simplified Geo-EAS format or a plain column file format. These formats are used in PCRaster for representation of point data, especially in the gstat module for geostatistical analysis. Point data column file format describes these formats. Point data are imported to PCRaster map format with the PCRaster operator col2map.
The data do not contain x and y coordinates: the asciifile with your data contains a sequence of cell values, without coordinates. In this case, the number of rows and columns of the PCRaster map must correspond exactly with the number of rows and columns of the file you want to import or else the import of data will result in nonsense. Among other data, maps from the ARC/INFO or Genamap GIS packages are imported in this way. It is done with the PCRaster operator asc2map.
Data can be exported from a PCRaster map using one of the operators map2col and map2asc.
The operator map2col exports data to an ascii column file in simplified Geo-EAS format or a plain column file format. Both contain x,y coordinates and data values. These kind of data are also used in the PCRaster package for representation of point data, see for formats Point data column file format. The plain column file format can easily be imported in spreadsheet, database management or word processing programs. It is also used in the gstat module of PCRaster.
The operator map2asc exports to an ascii file which will contain only data values, without x and y coordinates. This operator is used if you want to export data to the ARC/INFO package.
The spatial characteristics of the spatial data in PCRaster map format can be changed with both the operator resample and mapattr. Note that these two operations have a totally different result:
If you want to resample your data in a PCRaster map to a new map with different geographical location attributes use the operator resample. First create a new map with the location attributes you wish (this is done with mapattr). For instance it may have geographical location attributes that define an area that only partly covers the old map, with a somewhat smaller or larger cell size. Now the resample operator can be used to resample your old data to the new map: for each cell of the new map a new cell value is computed on basis of the cell values on the old map that come into the cell on the new map.
The resample operator can also be used to join two maps together. The maps that are joined together will be resampled and may have different location attributes: for instance they may overlap or may not overlap or may have different cell sizes.
The geographical location attributes of a map can be changed using the operator mapattr. Using this operator will not result in resampling of the data: each cell of the new map will contain a value that corresponds with the value on that cell of the old map. For instance halving the cell width of a map that consists of 50 x 50 cells of width 10 m. results in a (smaller) map of 50 x 50 cells of width 5 m., containing values that are taken directly from the old map. So changing the location attributes with mapattr will result in a new location of your data with respect to the real world coordinate system. Maybe this sounds silly but you may want to change the geographical location attributes after a map has been made, especially if you made an error in the specification of the location attributes with mapattr.
Conversion between data types is done using one of the conversion of data type operators (boolean, nominal, ordinal, scalar, directional and ldd. These operators change the data type of the input map to the data type that corresponds with the name of the operator. Conversion is only possible if it results in a map that has some meaning with the new data type attached to it.
hard copy output: not yet included in software
Visualisation of PCRaster maps and time series is done with aguila.
Creating and editing tables describes how to create or edit a table.
By default, the PCRaster package uses column tables. For defining relations between two maps of boolean, nominal, ordinal or ldd data type it is sometimes better to use matrices instead of column tables. If you want to use the matrix tables, set the global option --matrixtable. This option can be set for one separate operation or as general global option. How this is done, will be described later on in this chapter (Global options). If a relationship is specified between more than two maps in the matrix, the relation can not be described by a matrix table: PCRaster will automatically use a column table.
A table for the PCRaster operator lookup for creating a new map on basis of a table) or an input table for the operator table for counting the number of cells that match the key) can be made with a text editor or alternatively with spreadsheet or word processing programs by exporting your table as a file in ascii text format. Additionally, the table operator itself generates a correctly formatted table, which can be used (possibly with a few edits) as input table for the lookup operator.
The format of time series has been described in an earlier chapter (Time series format). Underneath (Creating and editing time series) database management with time series is described.
If you have time series data in a spread sheet program, database management program or a package for statistics you can create a PCRaster time series file by exporting your data as ascii formatted text. A PCRaster time series file must have the lay-out as described in Time series format. We advice to export the time series data in such a way that the resulting ascii formatted file will have a lay-out that looks like the lay-out used in PCRaster. Minor changes (for instance adding a header) can be made with a text editor.
Additionally, a timeseries file can be created by reporting a time series file to the database during a run of a dynamic model (see report keyword; and the timeoutput operator).
You can import a PCRaster time series file to an other software package by importing the time series as ascii text.
The format of point data column files has been explained in the previous chapter (Point data column file format). Underneath (Creating point data column files, conversion to/from PCRaster maps) the database management with point data column filess described.
If you have spatial data in a spread sheet program, database management program or a package for statistics you can create a point data column file by exporting your data as ascii formatted text. Before exporting, we advise you to put the x and y coordinates in the first and second column respectively. The third and following columns may contain the data. After exporting as ascii text, you can check and edit the ascii column file with a text editor. If you want to convert it to a PCRaster map it must have the simplified Geo-EAS format or the plain column file format, which is described in the previous chapter (Point data column file format). Conversion is done with the col2map operator (Creation of a PCRaster map; data import) about importing to a PCRaster map or see the operator col2map. Interpolation of regular or irregular spaced point data to a PCRaster map can be done with the gstat module.
Additionally, you can create a point data column file by exporting data from a PCRaster map, with the map2col operator. The point data column file will be in plain column file format or in simplified Geo-EAS format. See the map2col operator.
All applications have options to define the exact, detailed functionality of the operation. Two types of options are used in PCRaster. Local options begin with a single - (dash) are 1 character long and do mean some different depending on the application. Local options are documented in the application reference documentation.
Global options begin with a double - (dash) are 1 word long do have the same meaning across all applications.
If you want to set a different global option for only one operation,you can specify a global option in the command line. This is done in the same way as it is done for a local option, by typing the --globaloption after pcrcalc. If a non-pcrcalc operator is applied (for instance table) the global option is typed after the operator. The general global option setting described above is overruled only for the execution of the operation that is given in the command line.
In addition there are two other methods to set global options, who are explained next:
If you use the MS-Windows version of PCRaster, you can set general global options by typing after the command prompt:
set PCROPTIONS = globaloption1 globaloption2 ...globaloptionN
where globaloption1 is one of the global options, which, unlike a local option, is preceded by two - characters. If you use the UNIX version of PCRaster, global options are set by typing after the UNIX prompt:
PCROPTIONS=globaloption1 globaloption2 ...globaloptionNexport PCROPTIONS
Note that in UNIX no spaces are typed on either side of the = sign. For instance:
PCROPTIONS='--clone CloneStudyArea.map --lddin'; export PCROPTIONS
After applying PCROPTIONS the global options have the setting as specified or, if they are not specified, the default values. This set of options is used until a new set of options is specified with PCROPTIONS. If you set the options again with PCROPTIONS, options which are not specified that time are always set to default, no old settings are taken.
Global options that apply to a script can be stored in the script itself instead of typing it on the command line when the script is called. The first two characters of the first line of the script should read #!. After #! the global options can be specified. For example
#! --lddfill --radians
..etc.
Note
The #! should exactly be the first positions on the first line. Leading (empty) lines or spaces will change the line to a normal comment, and no global options from that line are set. No error messages are printed in such a case.
The remainder of this discussion on global options in the script applies only to those who use pcrcalc from a UNIX-like shell. The syntax of this first line is modelled after the alternative interpreter syntax of UNIX shells. If the first 2 characters of this first line are #! then all words on that line that can be recognized as global options, are parsed as global options, overriding global options at the command line. For example, example.mod:
#! /usr/local/bin/pcrcalc -F --lddfill
..etc.
To be fully compatible with the UNIX alternative interpreter notation, one needs to specify the -F flag on the first line if and only if one wants to put global options on the first line. Note that pcrcalc will always scan the first line of a script, even if the script is not executed by itself, in other words:
pcrcalc -f example.modis equal tochmod +x example.modexample.mod
In both cases the –lddfill option will be activated. Since pcrcalc scans the first line itself, these feature is also enabled on NON-UNIX OS’s (MSDOS).
-F is only neccessary and can only be used if one wants to specify global options. If the goal is to only describe pcrcalc as the interpreter of the script then use -f. This is due to a limitation in the syntax of an alternative interpreter, only one option can be specified. For example, example2.mod:
#! /usr/local/bin/pcrcalc -f
..etc.
global options related to location attributes:
global options for defining the sort of directional data type:
global options for defining cell representations:
These options define the representation of cell values used for storage and processing of data. The settings are only applied to the maps that are created: if you change the cell representation settings during a project, only the new maps you generate are assigned the new cell representation. The old maps created before you changed the settings will keep the ‘old’ cell representation.
scalar and directional data type: --single (default) or --double
Default, the cell representation is single real (--single). If you set --double it will result in double real representation of cell values. See also scalar data type, Scalar data type and directional data type, Directional data type.
nominal and ordinal data type: --small (default) / --large
Default, nominal and ordinal data are represented by smallninteger cell representation. If you set --large, large integer is used. See also and ordinal data type, Ordinal data type.
global option for specifying the format of tables
--columntable (default) or --matrixtable
Default, the operators lookup and table use column tables (setting --columntable). If the option --matrixtable is set, a matrix table is used. For a description of these formats, see Table format.
global options related to generation of a local drain direction map
--lddout (default) or --lddin
This option determines whether small catchments at the edge of the map are or are not considered as pits. See the operators lddcreate and lddcreatedem.
--lddfill (default) or --lddcut
This option determines the way the elevation model is modified in the corenof pits. See the operator lddcreatedem.
global option for specifying the neighbourhood of cells
--diagonal (default) or --nondiagonal This options specifies the neighbourhood of cells in the clump operation, see the clump operator.
global option for defining the sort of message printed during execution of an operation
--noprogress (default), --progress or --nothing
This option affects the message printed on the screen during execution of an operation. Default, the name, copyright and version of the software is printed and error messages if errors occur (--noprogress). The option --progress will result in additional information printed during execution, such as ‘busy with row x’, etc. If you set the option --nothing, nothing is printed on the screen, except error messages.
global option for defining the format of timeseries output. --noheader
Only use this option if the output format is not required to be compatible with PCRaster software components because the created timeseries can not be read back as a timeinput series nor can be vizualized by Aguila.
The option will affect timeseries as such that:
- No header is written.
- Selective reports for timeseries are enabled.
- no MV data is prepended when start time is larger than 1.
- Timeseries are only created at the end of the model run and not partial written when running the model. This means the -1 option is discarded.