kartograph.py is a Python script that renders shapefiles to SVG maps. The main idea is to get geo-referenced vector maps. For instance, s path that represents Argentina will store the ISO-code in a data-attribute:
<path data-iso="ARG" .. />
Also, the SVG files store information on the projection parameters, which can be used by the JS API to reconstruct the projection in order to add more geo-located layers and features (like markers).
Note that creating perfectly styled static maps is not the purpose of this script, so you won't find any options to set colors or styles. Instead, you will get really nice geo-references vector maps. They are intended to be rendered inside browsers where you can add lot's of CSS and JavaScript magic. Of course, you can open and edit them in Inkscape/Illustrator, too.
Kartograph.py is licensed under AGPL
kartograph.py is quite stable and renders really nice maps. Though, a few more things need to be done in the future:
- packaging as real Python module + pydocs
- multiple views per map (for instance to include Alaska and Hawaii to US map)
- and, of course, more map projections. Also see my wishlist.
The following Python libraries are required:
Also you need to download Natural Earth shapefiles following these download instructions
The following global options are avaiable
- --width, -w output width
- --height, -h output height
- --ratio, -r output ratio (will be used to compute missing width or height)
- --quality, -q quality level 0..100, see Quality section below
- --output, -o filename for the SVG map, if not provided tmp.svg will be used.
- --padding, -p how much spacing should be added around the map content
- --force-overwrite, -f by default, existing files will not be overwritten in batch mode, unless you set this parameter
- --list-projections prints a list of all available map projections
- --round-coordinates rounds all coordinates to integer values to reduce file size
Currently the the following commands are available
- world
- country
- regions
- region
- layer
- countries
- bbox
The command world renders a map of all countries in the world. The shapefiles come from the Natural Earth project.
For instance, this will output a world map into world.svg:
kartograph.py world --sea --width 600 --o world.svg
Also, you can now change the map projection and the center longitude:
kartograph.py world --sea --proj=robinson --lon0=-78
]
Command-specific options are:
- --sea, -s adds a sea background to the map
- --proj specify the map projection that should be used
- --lon0 sets the center longitude
- --filter filter out some countries
The country command renders maps that are centered on a country. You need to pass a three-letter country code (ISO 3166-1 alpha 3) as argument. For the country and regions maps the Lambert Azimuthal Equal-Area projection is used (again, more projections to come in future versions).
Usage:
kartograph.py country ISO
To render a map of Germany you would run
kartograph.py country DEU --height 300 -o DEU.svg
In the next example, the surrounding countries are added using the --context parameter. Also, the padding is set to 10% and the output ratio is set to 2 (width is two times height):
kartograph.py country DEU --h 300 --context --sea --padding 10 --ratio 2 -o DEU-context.svg
You can render several country maps at once by providing a comma-separated list of ISO-codes. In this case, only the directory name of the --output parameter will be used and the country maps are saved under their iso3 codes.
kartograph.py country DEU,FRA
Also, you can render country maps for all countries, although this may take a while to finish.
kartograph.py country all
Command-specify options are:
- --context, -c includes "surrounding" countries in the map to provide some context
- --context-quality use this if you want to set a different quality for the context (usually a lower quality)
- --sea will add a background indicating the sea
The regions (plural!) command works basically just like the country command, except for the selected country, the administrative-level-1 regions will be rendered instead of the country. The command provides the same parameters that are provided in the country command.
kartograph.py regions FRA --h 300 -o FRA-regions.svg
For instance, you can add context the same way as in country maps.
kartograph.py regions FRA --h 300 --context --sea --padding 10 --ratio 2 -o FRA-regions-context.svg
Command specific parameters:
- --join-regions join regions according to external region list (see below)
NOT IMPLEMENTED, YET
The region (singular!) command allows to render just a single region of a country.
kartograph.py region FRA --fips A4
Command specific parameters:
- --fips define the focus region by FIPS code
- --hasc define the focus region by HASC code
- --name define the focus region by name (will try to match minor spelling differences)
- --context adds surrounding regions and countries
- --region-context adds surrounding regions (no countries here)
The countries command renders a map that contains multiple countries and it's surrounding context. The following renders a map that (at least) contains Spain, Portugal, France, Germany and the United Kingdom, and the Lambert Conformal Conical projection (lcc) will be used.
kartograph.py countries ESP,PRT,FRA,DEU,GBR -p5 --proj=lcc -s
This can be used to add another layer to a SVG map. In general, every layer is represented by a SVG group (). The following example demonstrates the feature:
At first we create a country map of Brazil with some context.
kartograph.py country BRA --context --height 300 -p6 -s -o BRA.svg
Now we add a new layer for current forests, taken from Global Forest Watch.
kartograph.py layer BRA.svg globalforestwatch/w_curr.shp -o BRA-forests.svg
If we want just the forests in Brazil, we can use the --crop-to-layer parameter. We have to define the layer id to which the shapefile should be cropped to (in this case, the layer id is "BRA"):
kartograph.py layer BRA.svg globalforestwatch/w_curr.shp --crop-to-layer=BRA -o BRA-forests-cropped.svg
Here's another example: After rendering a country map of the United States we can add a layer for all counties spapes, taken from the Census 2000 shapefile.
kartograph.py country USA -o USA.svg
kartograph.py layer USA.svg census2000/co99_d00.shp -o USA-counties.svg
Command-specific parameters are:
- --layer-id will be used as id for the layer
- --crop-to-layer can be used to crop the shape to any existing layer
- --data-col here you can insert the shapefiles record column names that should be included in the SVG map (e.g. "ISO3,HASC")
The quality level will be used to compute the parameter for the polygon simplification (also called generalization). The higher the quality, the less the polygons are simplified. A quality of 100 means no simplification. Note that the overall quality also depends on the output size.
In some exceptional cases, the map data provided by Natural Earth admin-1 regions is too detailed. For instance, the following image shows the regions available for the United Kingdom.
Fortunately, in those cases the Natural Earth adm1 shapefile stores some values that identify the "parent" region. We can use this data to join regions. By now, there's a CSV file named data/region_joins.csv that stores a list of all regions that should be joined. You can activate region joining by setting the --join-regions (-j) parameter.
kartograph.py regions GBR --join-regions
The bbox command can be used to map custom lon/lat ranges.
kartograph bbox -19,-36,53,38 -g5 -s --proj stereo --lon0 16 --lat0 2 -l
