test suite needsdata to test udunits pkg about pcmdi_metrics HOT 22 CLOSED

How do 66Kb and 204Kb files grab you? I can provide tos and pr (pr was the variable that appeared to need the udunits call).

The tos file (GFDL) has units of "degC" and the obs (NOAA-OISST-v2) has units of "K".
The pr file (GFDL) has units of "kg/m2/s" and the obs (GPCP) has units of "kg m-2 s-1"

from pcmdi_metrics.

sounds perfectly reasonable

from pcmdi_metrics.

Oh and the files have absolutely no history or metadata.. Did you want me to update these with some info as to where they've been obtained from, when, who etc?

from pcmdi_metrics.

@doutriaux1 let me know if you need anything more else from me here - it would be a good idea to add another test with the issue described in #68 - this issue should be magically dealt with by the software, rather than the issue we're hitting.. And possibly throwing an error statement to let the user know they their data has problems (if that is the cause)

from pcmdi_metrics.

@doutriaux1 what do you need data-wise for this?

There are some GFDL-ESM2G files hanging off: oceanonly.llnl.gov slash durack1/dump/140808_metrics

from pcmdi_metrics.

@doutriaux1 what else do you need for this, is this closed via #145?

from pcmdi_metrics.

no we do not test for wrong/different units in this case.
I need to basically add 2 test case one for this and one for tas as @gleckler1 requested.

from pcmdi_metrics.

@doutriaux1 both TAS and TOS fields with the GFDL units (degrees_C) are findable at the URL (broken) above, and repeated here: oceanonly.llnl.gov slash durack1/dump/140808_metrics

from pcmdi_metrics.

@doutriaux1 this is 1 of 4 open issues which are associated with the 1.0 milestone.. Any work required on this or is it closeable?

from pcmdi_metrics.

let me see if I could reuse some of the example data we have via a cdscan and alter units even though it would lead to wrong result, we could check that we indeed get the wrong results, hence udunits worked

from pcmdi_metrics.

@doutriaux1 I don't think you need too.. The GFDL-ESM2G files that are findable at the links above should trigger UDUNITS..

from pcmdi_metrics.

how big are they. i'm trying to keep the repo small

from pcmdi_metrics.

66KB and 204KB as it says in #52 (comment)

from pcmdi_metrics.

I should learn to read....

from pcmdi_metrics.

I do try to be overly verbose in my comments.. You should know this by now..

from pcmdi_metrics.

though I've just noted that I gave you TAS and TOS, and not PR (which as my verbose comment noted: #52 (comment)) this is the variable which appeared to trigger udunits.. Want PR too?

from pcmdi_metrics.

path above is dead. Which version has them?

from pcmdi_metrics.

really need to learn to read it does say http:// ....

from pcmdi_metrics.

that's the reason why the path wasn't written complete above.. so that google scours don't see this..

from pcmdi_metrics.

ok i'm almsot setup... but i do need 12 months of data your seem to have 1 only

from pcmdi_metrics.

if you want a full suite take a peek at /export/durack1/140701_metrics/test and you should have all the variables I've been testing..

from pcmdi_metrics.

no worries got a working test with existing datasets and cdscan

from pcmdi_metrics.