Re: [netcdfgroup] Scipy netcdf issue

Hi Roy, Wei-keng,


On Fri, Jun 22, 2018 19:18:42 -0500, Wei-keng Liao wrote:
> The file is corrupted. I ran it against ncvalidator and got the following 
> errors.

Thanks for looking into this. Someone had commented on the scipy issue[1]
about the offsets but I hadn't understood completely what they'd meant.
Thanks for clarifying it up.

Would the suggested way of handling such files be to simply flag them as
corrupt and not process them at all then?

While the scipy implementation crashes with this file, the netcdf
implementation that nibabel uses does not. I'm still trying to figure
out what they do differently.

[1] https://github.com/scipy/scipy/issues/8153#issuecomment-364757198

-- 
Thanks,
Regards,

Ankur Sinha "FranciscoD"

https://fedoraproject.org/wiki/User:Ankursinha

Attachment: signature.asc
Description: PGP signature