bring pvlib-python up to date with PVLIB-MATLAB 1.2 and 1.3

[robwandrews]

Add additional functions introduces in PVLIB-matlab 1.2

[dacoex]

Where can we find PVLIB-matlab 1.2?
Is it this one?
https://github.com/Sandia-Labs/PVLIB_Matlab/commits/master

Is the documentation of functions online?
It looks that one needs to be logged in:
https://pvpmc.sandia.gov/resources-for-members/pv_lib-toolbox/matlab/functions-by-catagory/

And is it planned to develop matlab & python libs always in parallel?

anyway, I would like to see in future Shading and Soiling (Item 3) as well as AC Losses (Item 9) of
https://pvpmc.sandia.gov/wp-content/uploads/2014/08/1199X499_slide2.png
But maybe these deserve a separate page/issue.

[wholmgren]

PVLIB-Matlab 1.2 is only available at the PVPMC website. The github PVLIB_Matlab repository is an older version and may be removed. @Calama-Consulting you might want to ask Josh if he wants the matlab project available on github since it may hurt their official download statistics.

The readme has a link to the documentation.

We want to keep the core functionality and algorithms of the python and matlab projects roughly in parallel, but we're not making any promises at this point.

Yes, please do make separate issues for any feature requests.

[dacoex]

PVLIB-Matlab 1.2 is only available at the PVPMC website. The github PVLIB_Matlab repository is
an older version and may be removed.

OK, so login and find out...

The readme has a link to the documentation.

Which is behind a login gate.

We want to keep the core functionality and algorithms of the python and matlab projects roughly in >parallel, but we're not making any promises at this point.

So this means:
PVPMC adds a functionality and then issues a release, then the pvlib-python would be updated?
A rather parallel process?
In consequence, one would have to know Mtalab, as well?

Yes, please do make separate issues for any feature requests.

OK, one the development strategy discussed here is clear.

[wholmgren]

The read me for this github repo has a line

"Full documentation can be found at readthedocs"

Which links to

http://pvlib-python.readthedocs.org/en/latest/

This is the Pvlib Python documentation only.

The Matlab and Python projects are developed by different teams that do not
regularly work together. There will not be feature parity, only similarity.
On Tue, Feb 17, 2015 at 2:00 PM Data & Code Expert Experimenting with Code
on Data notifications@github.com wrote:

PVLIB-Matlab 1.2 is only available at the PVPMC website. The github
PVLIB_Matlab repository is
an older version and may be removed.

OK, so login and find out...

The readme has a link to the documentation.

Which is behind a login gate.

We want to keep the core functionality and algorithms of the python and
matlab projects roughly in >parallel, but we're not making any promises at
this point.

So this means:
PVPMC adds a functionality and then issues a release, then the
pvlib-python would be updated?
A rather parallel process?
In consequence, one would have to know Mtalab, as well?

Yes, please do make separate issues for any feature requests.

OK, one the development strategy discussed here is clear.

—
Reply to this email directly or view it on GitHub
#2 (comment).

[dacoex]

The Matlab and Python projects are developed by different teams that do not regularly work together. There will not be feature parity, only similarity.

OK, this is more clear.
Consequently, you may decide to spearhead with new features and then the Matlab will follow?

Or is pvlib after a similar feature set but more driven by its developers/users?

Maybe good also to limit the scope and compare it to other packages such as https://github.com/pingswept/pysolar. Then teh docs cloud show where pvlib by its intention will be different apart from the more sphisticated tied-in to the scipy stack with the reliance on pandas (very positive).

I am thinking of something like:
Related Packages
http://statsmodels.sourceforge.net/devel/related.html

Looking forward to your feedback on this and the above questions.

[wholmgren]

Pvlib Python already has a few features that the Matlab version doesn't
have. It's up to the Matlab developers to update their code where possible.
I can only say that the Python project will try to stay current with most
things that the Matlab project adds.

The Python project will be mostly community supported, whereas the Matlab
project is mostly supported by u.s. department of energy grants.

The Python project is very much tied to the scientific Python stack. This
would not be feasible without those libraries.

Our pvsc 2015 abstract and rob's 2014 PVSC paper outline the scope of this
project.

Pysolar is much more limited in scope and is gpl licensed. Nothing against
Pysolar, it's just different.

I will put some of this in the docs/readme.
On Tue, Feb 17, 2015 at 3:45 PM Data & Code Expert Experimenting with Code
on Data notifications@github.com wrote:

The Matlab and Python projects are developed by different teams that do
not regularly work together. There will not be feature parity, only
similarity.

OK, this is more clear.
Consequently, you may decide to spearhead with new features and then the
Matlab will follow?

Or is pvlib after a similar feature set but more driven by its
developers/users?

Maybe good also to limit the scope and compare it to other packages such
as https://github.com/pingswept/pysolar. Then teh docs cloud show where
pvlib by its intention will be different apart from the more sphisticated
tied-in to the scipy stack with the reliance on pandas (very positive).

I am thinking of something like:
Related Packages
http://statsmodels.sourceforge.net/devel/related.html

Looking forward to your feedback on this and the above questions.

—
Reply to this email directly or view it on GitHub
#2 (comment).

[dacoex]

Thanks a lot.

So next could be a series of feature plans that can be added to the roadmap.

[wholmgren]

Here's a list of the v1.2 Matlab project's functions that we're missing (from the 1.2 help page):

• pvl_louche - Estimate Direct Normal Irradiance from Global Horizontal Irradiance using the LOUCHE model
• pvl_orgill_hollands - Estimate Direct Normal Irradiance from Global Horizontal Irradiance using the Orgill and Hollands model
• pvl_erbs - Estimate Direct Normal Irradiance from Global Horizontal Irradiance using the Erbs model
• pvl_reindl_1 - Estimate Direct Normal Irradiance from Global Horizontal Irradiance using the Reindl 1 model
• pvl_reindl_2 - Estimate Direct Normal Irradiance from Global Horizontal Irradiance using the Reindl 2 model
• pvl_martinruiziam - Determine the incidence angle modifier using the Martin and Ruiz incident angle model
• pvl_adrinverter - Anton Driesse's Grid-Connected PV Inverter

I also looked at the changelog but most of it has either already been incorporated or is Matlab specific.

Many of these would be good first PRs.

[bmu]

Maybe Anton could implement "his Inverter" ;-) @adriesse?