Importing data: Difference between revisions

From GRASS-Wiki
Jump to navigation Jump to search
m (+category)
(→‎Modules: r.in.gdal memory hint)
Line 73: Line 73:
|}
|}
</center>
</center>
=== Large data ===
When importing large raster maps with {{cmd|r.in.gdal}} the default internal settings may be slow, with a large percentage of the CPU locked up in kernel system operations. In these cases you should increase the ''memory='' option to be larger than:
tile_height * raster_width * band_count * sizeof_datatype
You can find those values with <tt>gdalinfo</tt>.
Setting the ''r.in.gdal memory='' option has the same effect as setting the GDAL_CACHEMAX environment variable.


=== Modules ===
=== Modules ===

Revision as of 00:01, 24 January 2014

Getting your data into GRASS

GRASS supports many common GIS formats, and many uncommon ones too.

The primary import/export tool is called GDAL/OGR. It is the Rosetta Stone of the free GIS world:


Create a new location for your data with the correct map projection using the Location Wizard on the main startup screen^, then in the menus:

Raster formats:  File → Import raster data using GDAL
Vector formats:  File → Import vector data using OGR

Typically GDAL/OGR will automatically detect the file format for you.

  • [^] You must create a new GRASS location for each map projection you will work with or import data from. If you want the data reprojected into something else you'll have to create that target map projection in a location and pull them into there with r.proj and v.proj, or use GDAL's gdalwarp or ogr2ogr command line tools.
[read more about this on the projectionintro help page]


  • Many GPS formats can imported with GPSBabel as well using the v.in.gpsbabel tool.
  • CSV files from a spreadsheet can be imported with the v.in.ascii module.
    (v.in.ogr could too, but it's not as smooth)
    • If your .csv file contains "quoted" strings containing commas, you can use the csv_dequote.pl script to parse them into a less-problematic form ready for import with v.in.ascii.


After the map has been imported you can add it to the display layer list in the Layer Manager window. You will most likely want to right click on the map in the layer list and select "Zoom to selected map(s)". Otherwise when you go to view the map you could be far away and just see a white display canvas.

To view a raster map, add it to the Layer Manager with the grid-like button with a (+), to add a vector map to the Layer Manager use the button with lines and points on it to the right of the add raster map button. The left hand button in the Map Display window will re-render the map if needed.

The "computational region" is a very important concept in GRASS and is distinct from the "display window" you see in the Map Display window. You can also read about it in the rasterintro help page. You can set the computational region and cell resolution to match a specific map with the g.region module.

After importing your data you need to run "g.region -p rast=your_map" or "g.region -p vect=your_map" to set the region bounds/zoom to match that of the newly imported map. Be careful with the region resolution when doing this with vector maps, you may have to run "g.region -a res=" to clean that up and keep it reasonable (for example 2000x2000 rows and columns).

Import Video Tutorials

External links

You can just link to the data rather than importing them.

  • r.external: full support for external raster data (all GDAL supported formats)
  • v.external: basic support for external vector data with some topology limitations (all OGR supported formats)

Bulk data import

The wxGUI menu tools offer simplified import.

Importing multiple vector (e.g., SHAPE) files

Bulk import of vector files via menu
Bulk import of multiple SHAPES via menu (files are auto-selected)

Importing multiple raster (e.g., GeoTIFF) files

Bulk import of GeoTIFF via menu
Bulk import of GeoTIFF via menu (files are auto-selected)

Large data

When importing large raster maps with r.in.gdal the default internal settings may be slow, with a large percentage of the CPU locked up in kernel system operations. In these cases you should increase the memory= option to be larger than:

tile_height * raster_width * band_count * sizeof_datatype

You can find those values with gdalinfo.

Setting the r.in.gdal memory= option has the same effect as setting the GDAL_CACHEMAX environment variable.

Modules

Raster

Vector and points

Data formats