SRTM GeoTIFF Import - GAMA shows "Unknown Format"

98 views
Skip to first unread message

Sonja Holler

unread,
Jan 2, 2020, 7:26:04 AM1/2/20
to GAMA
Dear all,

I am trying to use a SRTM 1km clip for High Asia (clipped and exported to GeoTIFF with QGIS, 200MB), but GAMA doesn't recognize it (the "includes" folder shows ... .tif (-1x-1| Unknown Format)). I tested the files with the small geotiff import model, but GAMA cannot produce the grid/map, the bogota_grid.tif file works though.

I tried the QGIS export options 1) rawdata, 2) as displayed 3) VRT as variations but the result remains the same.

I'm using GAMA V1.8.0 on Mac.

Can anybody tell me how to fix this?

Thank you very much in advance!

Best regards, Sonja

Huynh Quang Nghi 002628

unread,
Jan 2, 2020, 7:31:48 AM1/2/20
to gama-p...@googlegroups.com
Hi, could you please try the newest build
We had an issue liie you said in navigator and it seems be fixed recently.
However, did your data work when you import and simulate the model?

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/2d59d613-34fe-43d4-a302-b112aa27decf%40googlegroups.com.

Sonja Holler

unread,
Jan 2, 2020, 9:16:32 AM1/2/20
to gama-p...@googlegroups.com
Hi,
I tried the newest build and GAMA is now able to import the file and to show its properties in the "includes" folder (5903x4237 | TIFF). Thank you, thats one step forward!

However, I do not get GAMA to display it. I already allocated 32GB of RAM to GAMA, but it will still run out of memory or crash when I try to display the geotiff (200MB) with the geotiff import model.

Any Ideas what I do wrong or what could be wrong with the file? Or is GAMA not able to handle that many cells?

Thank you for helping!

Best regards, Sonja

To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/CANNWKocNXW1wOiNcFaX8DHw8yYx-s9sX5byoKyXZK6uBnoCgbg%40mail.gmail.com.

Huynh Quang Nghi 002628

unread,
Jan 2, 2020, 9:30:24 AM1/2/20
to gama-p...@googlegroups.com
Hi,
If i understand correctly, your data will create approx 25 millions agent in gama. i dont know if 32gb ram is enough or not, and if the memory is not the problem, it would also take long time in computation after that.
SO , Let s say YES, GAMA cant handle such  huge amount of cells. You should decrease the size. or .... someone else please help.

To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/CADqHXRq2H8hSMiud2TVe9sCfz6WHj%2B%2BMbZPeRH8NrDtyke85eg%40mail.gmail.com.


--
-------
HUYNH Quang-Nghi, Ph.D.
Lecturer
College of Information and Communication Technology - Can Tho University
Campus II, 3/2 street, Ninh Kieu district
Can Tho city, Vietnam
[Vietnam] +84909128322

Sonja Holler

unread,
Jan 3, 2020, 3:31:54 AM1/3/20
to gama-p...@googlegroups.com
Ok,

thank you for your reply, I may have to approach this yet another way to create less volume.

Best regards, Sonja

To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/CANNWKodEZc3Uq7BEen%2BqSVz_3AzUrBPe23g5QnJ_0kx64_MR2Q%40mail.gmail.com.

Srirama Bhamidipati

unread,
Jan 3, 2020, 3:53:47 AM1/3/20
to GAMA
Hi

what is the final use of your TIFF in GAMA? does it have any dynamic behavior during the simulation or acts just as a background or is it just to create initial condition?

Srirama 
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.


--
-------
HUYNH Quang-Nghi, Ph.D.
Lecturer
College of Information and Communication Technology - Can Tho University
Campus II, 3/2 street, Ninh Kieu district
Can Tho city, Vietnam
[Vietnam] +84909128322

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

Sonja Holler

unread,
Jan 3, 2020, 8:15:29 AM1/3/20
to gama-p...@googlegroups.com
Hi Srirama,

i want to use altitude as an initial condition to place the agents in their habitat and during the simulation for movement restrictions. I initially wanted to display the global population but the habitatarea seems to be to big to handle for GAMA.

I managed to incorporate the size of Bhutan now and 150 additional agents (also it takes time loading), but my code seems to be confusing for GAMA, since agents are placed outside the targeted altitudes and I get now a lot of nil errors, which seem to be based on the grid movement of accessible neighborhood cells based on the altitude restriction. Also, some of my agents jump all over the place now, in stead of choosing neighborhood cells.

I try to figure this out.

If you have still an idea how to incorporate an 1km SRTM with 5903x4237 cells and 4000 to 10.000 additional agents it would be very helpful ...

best,

Sonja

To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.


--
-------
HUYNH Quang-Nghi, Ph.D.
Lecturer
College of Information and Communication Technology - Can Tho University
Campus II, 3/2 street, Ninh Kieu district
Can Tho city, Vietnam
[Vietnam] +84909128322

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/6d99dafb-3e9f-4974-8e33-a582ec10e507%40googlegroups.com.

Srirama Bhamidipati

unread,
Jan 9, 2020, 5:08:00 AM1/9/20
to gama-p...@googlegroups.com
Hi,

I think a million agents would indeed be too many. But still I usually approach it using a vector mesh instead of a raster grid. 

Or perhaps even triangulate  to a smoother surface so that the number of cells/ triangles are drastically reduced without compromising on the surface of interest. 
You can also "dissolve" cells of similar properties to reduce the number of cells. 
You could also consider multiple "subsets" of your model as shown in the library "https://gama-platform.github.io/wiki/RunSeveralSimulations"
If you are working on movements, you can also "skeletonize" a probable path for these movements, I think there is an example on people moving withing a building, you can adapt this to your terrain. 
Or you can consider it like a "reinforcement learning" approach ... run a training procedure (intense) ... but then use  the result for potential "trained" movements to reduce the exploration space for movements. 

Are you working on Bhutan case? very interesting :) beautiful country super nice people ! 

Srirama

To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.


--
-------
HUYNH Quang-Nghi, Ph.D.
Lecturer
College of Information and Communication Technology - Can Tho University
Campus II, 3/2 street, Ninh Kieu district
Can Tho city, Vietnam
[Vietnam] +84909128322

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platform+unsubscribe@googlegroups.com.

Sonja Holler

unread,
Jan 9, 2020, 2:51:11 PM1/9/20
to gama-p...@googlegroups.com
Hi Srirama,

I have a probability raster as the food source for my agents, so I will stick to the raster grid, but thank you very much for your effort, I will look into "subsets"!

Best regards,

Sonja

Am Do., 9. Jan. 2020 um 11:08 Uhr schrieb Srirama Bhamidipati <b.sr...@gmail.com>:
Hi,

I think a million agents would indeed be too many. But still I usually approach it using a vector mesh instead of a raster grid. 

Or perhaps even triangulate  to a smoother surface so that the number of cells/ triangles are drastically reduced without compromising on the surface of interest. 
You can also "dissolve" cells of similar properties to reduce the number of cells. 
You could also consider multiple "subsets" of your model as shown in the library "https://gama-platform.github.io/wiki/RunSeveralSimulations"

To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.


--
-------
HUYNH Quang-Nghi, Ph.D.
Lecturer
College of Information and Communication Technology - Can Tho University
Campus II, 3/2 street, Ninh Kieu district
Can Tho city, Vietnam
[Vietnam] +84909128322

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "GAMA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gama-platfor...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gama-platform/625bbe8b-7e33-4b31-b0b1-4125c971f42f%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages