[Neuroimaging] Bumping up minimal numpy version?

[Matthew Brett]

Hi,

On Tue, Oct 13, 2015 at 1:14 PM, Ariel Rokem <arokem at gmail.com> wrote:
>
>
> On Tue, Oct 13, 2015 at 12:42 PM, Matthew Brett <matthew.brett at gmail.com>
> wrote:
>>
>> Yo,
>>
>> On Tue, Oct 13, 2015 at 12:40 PM, Ariel Rokem <arokem at gmail.com> wrote:
>> > Hi Matthew,
>> >
>> > Thanks for the quick response!
>> >
>> > On Tue, Oct 13, 2015 at 12:27 PM, Matthew Brett
>> > <matthew.brett at gmail.com>
>> > wrote:
>> >>
>> >> Hi,
>> >>
>> >> On Tue, Oct 13, 2015 at 12:18 PM, Ariel Rokem <arokem at gmail.com> wrote:
>> >> > Hi everyone,
>> >> >
>> >> > With numpy 1.10 now out, would it be draconic to bump up the minimal
>> >> > supported version up to 1.8?
>> >> >
>> >> > See also: https://github.com/nipy/dipy/pull/731
>> >>
>> >> Can you avoid that?  Standard minimum numpy requirement seems to be
>> >> 1.7.1 - e.g Pandas, Scipy etc.
>> >
>> >
>> > Probably. But there doesn't seem to be a wheel for that in the
>> > scikit-image
>> > wheelhouse:
>> >
>> > https://travis-ci.org/nipy/dipy/jobs/85177837#L171
>> >
>> > Would it be much trouble to make that available?
>>
>> Not much trouble, but there is a wheel for 1.7.1 .  I can't remember
>> what the 1.7.1 was compared to 1.7.0 but it does seem to be standard
>> to use 1.7.1.
>
>
> Tried it, but I really don't think there's a wheel for 1.7.1:
> https://travis-ci.org/arokem/dipy/jobs/85194895#L164

Oops - I see that I didn't build one for Python 2.7 ...  I'll do that now.

Matthew