Eigen Include

[Michael Betancourt]

What’s the current requirement for including Eigen in Stan code (just Eigen, nothing else)?

[Bob Carpenter]

I'm pretty sure we only use the main Eigen library in compiling
Stan programs. bin/print uses the unsupported FFT module for
autocorrelation in n_eff calculations.

- Bob

> On Jun 23, 2015, at 7:56 AM, Michael Betancourt <betan...@gmail.com> wrote:
>
> What’s the current requirement for including Eigen in Stan code (just Eigen, nothing else)?
>

> --
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
>

[Michael Betancourt]

I mean the minimal include — before including <Eigen/Core> directly would
break all of the metaprograms so there was a special order in which things
had to be included. With the refactor of math what is the current state of
using Eigen in internal code?

[Ben Goodrich]

I think it is just

#include<stan/math/prim/mat/fun/Eigen.hpp>

which seems to just come alphabetically in math.hpp.

Ben

On Tuesday, June 23, 2015 at 12:59:19 PM UTC-4, Michael Betancourt wrote:

I mean the minimal include — before including <Eigen/Core> directly would
break all of the metaprograms so there was a special order in which things
had to be included.  With the refactor of math what is the current state of
using Eigen in internal code?

On Jun 23, 2015, at 5:50 PM, Bob Carpenter <ca...@alias-i.com> wrote:

> I'm pretty sure we only use the main Eigen library in compiling
> Stan programs.  bin/print uses the unsupported FFT module for
> autocorrelation in n_eff calculations.
>
> - Bob
>
>> On Jun 23, 2015, at 7:56 AM, Michael Betancourt <betan...@gmail.com> wrote:
>>
>> What’s the current requirement for including Eigen in Stan code (just Eigen, nothing else)?
>>
>> --
>> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.

>> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+unsubscribe@googlegroups.com.

>> For more options, visit https://groups.google.com/d/optout.
>>
>>
>
> --
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+unsubscribe@googlegroups.com.

[Bob Carpenter]

Correct --- I refactored the order dependency away by
removing our code injection of a typedef into Eigen and replacing
it with a template metaprogram to infer scalar types of a container.

- Bob

> On Jun 23, 2015, at 1:05 PM, Ben Goodrich <goodri...@gmail.com> wrote:
>
> I think it is just
>
> #include<stan/math/prim/mat/fun/Eigen.hpp>
>
> which seems to just come alphabetically in math.hpp.
>
> Ben
>
> On Tuesday, June 23, 2015 at 12:59:19 PM UTC-4, Michael Betancourt wrote:
> I mean the minimal include — before including <Eigen/Core> directly would
> break all of the metaprograms so there was a special order in which things
> had to be included. With the refactor of math what is the current state of
> using Eigen in internal code?
>
> On Jun 23, 2015, at 5:50 PM, Bob Carpenter <ca...@alias-i.com> wrote:
>
> > I'm pretty sure we only use the main Eigen library in compiling
> > Stan programs. bin/print uses the unsupported FFT module for
> > autocorrelation in n_eff calculations.
> >
> > - Bob
> >
> >> On Jun 23, 2015, at 7:56 AM, Michael Betancourt <betan...@gmail.com> wrote:
> >>
> >> What’s the current requirement for including Eigen in Stan code (just Eigen, nothing else)?
> >>
> >> --
> >> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.

> >> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.

> >> For more options, visit https://groups.google.com/d/optout.
> >>
> >>
> >
> > --
> > You received this message because you are subscribed to the Google Groups "stan development mailing list" group.

> > To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.

> > For more options, visit https://groups.google.com/d/optout.
>
>
> --
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.