Impact function for generic categorized raster hazard no longer exist? #531

Closed
vdeparday opened this Issue Mar 11, 2013 · 10 comments

Projects

None yet

4 participants

@vdeparday

In 1.1, I used a function for categorized raster hazard and building, my layer were tagged like this:
{'category': 'hazard', 'subcategory': 'N/A', 'unit': 'normalised', 'title': 'sri_lanka_tsunami_hazard'}
{'category': 'exposure', 'subcategory': 'structure', 'title': 'Building_SIIRM'}

But now, in the master version, with the same dataset, the function is no longer available.

Were there some clean-up in the function? the tagging has changed?

@ismailsunni ismailsunni was assigned Mar 11, 2013
@timlinux

Would you mind making the datasets available (e.g. via dropbox)? @ismailsunni could you take a look?

Thanks!

Tim

@vdeparday

You can use the qgs project available in this DropBox folder. There are vector and raster data but only the raster data is relevant for the function described:
https://www.dropbox.com/sh/czknfdf0c5oas0e/hBOnigouvL

@ismailsunni
InaSAFE member

Hi all

I think @uniomni move them to test-only area in 9696df7

@ismailsunni
InaSAFE member

I have talked to @uniomni , and we agree to put it back in InaSAFE so that you can use it again.

Wait for it

@ismailsunni
InaSAFE member

I have put the categorized impact function for population back. But, not for the categorized impact function for building, since I found some mistake in the impact function. So, i will keep it in the test-only directory until we fix it.

@cchristelis cchristelis was assigned Sep 2, 2013
@timlinux

@cchristelis can you take this ticket over since Ismail has left the project?

Thanks

Tim

@cchristelis

I will have a look at this. I'd like to know what the mistake is that ismailsunni found in the impact function though...

@ismailsunni
InaSAFE member

Hi @cchristelis , I'm so sorry, but unfortunately, I forget what's the mistake. May be you can just make it available and see what's the mistake. I just read the code and I can't remember what is it.

@cchristelis

This issue is resolved by:

#1024

@cchristelis

The impact function has a problem close to zone edges. This seems to be due to interpolation. The rounded number is compared to the int value and as such falsely calls a building unaffected.

@cchristelis cchristelis added the Ready label Jan 7, 2015
@cchristelis cchristelis closed this Feb 9, 2015
@cchristelis cchristelis removed the Ready label Feb 9, 2015
@cchristelis cchristelis was unassigned by vdeparday Nov 16, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment