30 day month?

[tommibergman]

Hi, I am using version 1.7 from Conda to aggregate AERONET collocated data to monthly data.
My problem is that I get 13 months since CIS is using 30 day month.

So using:
cis aggregate od550aer:../output/processed/col_aeronet_oldsoa-bhn-2010_bak/all/SACOL.lev20.2010.nc t=[2010-01-01,2010-12-31,P1M] -o test.nc

I get times (from ncdump):

time = 149765, 149795, 149825, 149855, 149885, 149915, 149945, 149975, 
    150005, 150035, 150065, 150095, 150125 ;

How I can make it go with variable month length instead? Or is this a bug? I used earlier version 1.4 and I did not have this problem.

[duncanwp]

Yes, this should be calendar months. I’ll take a look and let you know. Are you sure this worked in 1.4 though? On 1 Jul 2019, at 14:43, Tommi Bergman <notifications@github.com<mailto:notifications@github.com>> wrote: Hi, I am using version 1.7 from Conda to aggregate AERONET collocated data to monthly data. My problem is that I get 13 months since CIS is using 30 day month. How I can make it go with variable month length instead? Or is this a bug? I used earlier version 1.4 and I did not have this problem. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#20?email_source=notifications&email_token=AAYF2VDW3UCWPYMPOMB75M3P5ICSJA5CNFSM4H4SQ2C2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4G4USK5A>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AAYF2VAIAPIDGNFZFTDCAT3P5ICSJANCNFSM4H4SQ2CQ>.

[tommibergman]

Yes, my old aggregated files created using 1.4.0 have 12 months and separation of time stamps correspond to actual month lengths.