[GRASS-dev] GSOC project proposal
wenzeslaus at gmail.com
Mon Mar 9 20:28:02 PDT 2015
On Mon, Mar 9, 2015 at 6:32 PM, sarthak agarwal <sarthak0415 at gmail.com>
> Dear Sir/Madam,
> Myself Sarthak currently working under Dr. KS Rajan as a research student
> in Spatial Informatics Lab, IIIT hyderabad . I am currently working in
> spatial databases by optimizing the performance of mongoDb for spatial
> databases. Previously i have worked on Billboard placement optimization
> problem considering the spatial ,temporal and traffic features.
> I am interested in working with grass org under the project "New
> easy-to-use command line interface for GRASS GIS*". *The problem looks
> interesting to me. I am familiar with the technologies needed for this
> project.As i have already worked in c++ and python .
this sounds good. You should find some enhancement or bug in GRASS GIS bug
tracker  and implement or fix before the application evaluation it to
show level of you proficiency [2, 3].
For example an enhancement ticket #1204 related to invoking GRASS. The wish
is to start grass7 command with a wxGUI workspace file as a parameter .
Another option is to continue in implementing of related #2579, enhancement
to provide command line parameter to execute modules. However, the big
issue there is more the discussion about the interface rather then the
implementation itself which is partially done already.
More challenging option is to implement some part of the GSoC project, e.g.
for rasters or for PostGIS vectors. This is of course highly related task
but it can be tricky and what you need is something which is working well
before the application evaluation. More smaller tasks might a better idea.
To understand how projections (coordinate systems) work in GRASS GIS (see
below), it might be advantageous for you to implement r3.proj module which
would project 3D rasters from one GRASS Location to another. r.proj module
is in C but for start r3.proj module could be in Python and would use
r.proj in the background (similar idea can be ussed for GRASS imagery
groups and temporal datasets which also don't have their "proj" module).
Further, you could try to write r3.in.proj module which would import 3D
raster data in one projection to GRASS Database with another projection
(see source code of r.in.proj). See below for more resources on this topic.
Writing tests for related things such as #2579, or modules r.in.proj,
v.in.proj or the other modules mentioned below might be quite good
introduction to the topic but to show your skill you might want to combine
it with some other task which would involve actual coding.
However, you can explore the open issues on GRASS GIS bug tracker yourself.
Just keep in mind that this GSoC project would be in Python, so the thing
you select be in Python.
In any case, start with compiling GRASS GIS from source code from
Subversion repository .
> Can you please provide me with some more literature and links to
> understand the underlying concept behind the problem and a direction to
> solve it.
Start with running GRASS GIS and playing with data in different projections
in case you are not familiar with GRASS GIS already. Read through the
related manual page  and/or tutorials online on getting your data into
Definitively read the code from various projects and other things linked in
the GSoC idea . Go through the modules related to importing, linking,
exporting and projecting data, some of them are linked below. Look also at
how GRASS GIS can connect to PostGIS .
Source code for r.in.proj and v.in.proj:
Also I want to know whether someone is already committed to this Project or
> No, at least nobody wrote to the mailing list.
> Sarthak Agarwal
> Btech and Ms by research in Spatial Informatics
> International Institute Of Information Technology ,Hyderabad
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the grass-dev