[Draft] add single-argument kernel evaluator

[lu1and10]

To address issue #528, added a single-argument kernel evaluator using current polynomial coefficients.

@ahbarnett In case you want to test the coefficients sooner.
I tested by replacing evaluate_kernel((FLT)z[n], opts); with evaluate_kernel_horner((FLT)z[n], opts); in functions onedim_fseries_kernel and onedim_nuft_kernel, make test still passes.

This may be merged after 2.3 release.

[ahbarnett]

Excellent! - I look forward to trying this out. Glad to know the switch works. I would like to set it up so kerevalmeth=0,1, switches *both* spreadinterp and finufft:onedim_*_kernel together, so they always use the same kernel, be it ES or Horner. Opinions how to do that in cleanest way welcome.. Thanks, Alex

…

On Wed, Aug 28, 2024 at 7:21 AM Libin Lu ***@***.***> wrote: To address issue #528 <#528>, added a single-argument kernel evaluator using current polynomial coefficients. @ahbarnett <https://github.com/ahbarnett> In case you want to test the coefficients sooner. I tested by replacing evaluate_kernel((FLT)z[n], opts); with evaluate_kernel_horner((FLT)z[n], opts); in functions onedim_fseries_kernel and onedim_nuft_kernel, make test still passes. This may be merged after 2.3 release. ------------------------------ You can view, comment on, or merge this pull request online at: #541 Commit Summary - aa2be60 <aa2be60> add single-argument kernel evaluator File Changes (2 files <https://github.com/flatironinstitute/finufft/pull/541/files>) - *M* include/finufft/spreadinterp.h <https://github.com/flatironinstitute/finufft/pull/541/files#diff-4a0efeca16f9348de45eace7da9efbdddfa20170009474275fb861d4780bb0cb> (2) - *M* src/spreadinterp.cpp <https://github.com/flatironinstitute/finufft/pull/541/files#diff-183554bb1a791ce2cfe2a1149dffebabd6bddb30b97ec42e4da0fe823fa44802> (76) Patch Links: - https://github.com/flatironinstitute/finufft/pull/541.patch - https://github.com/flatironinstitute/finufft/pull/541.diff — Reply to this email directly, view it on GitHub <#541>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACNZRSXS4PW4QT6TPIA2QK3ZTWXD7AVCNFSM6AAAAABNIB4UP2VHI2DSMVQWIX3LMV43ASLTON2WKOZSGQ4TCOBQGU4DKNA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

-- *-------------------------------------------------------------------~^`^~._.~' |\ Alex Barnett Center for Computational Mathematics, Flatiron Institute | \ http://users.flatironinstitute.org/~ahb 646-876-5942

[DiamonDinoia]

Maybe throwing an exception is not the most evil thing as the user cannot reach throw since this is internal API.

[DiamonDinoia]

Is this ready to merge? Can you write a test that checks that both code (even odd and this) give same result?

[mreineck]

This PR will almist certainly conflict with #567. If #567 is merged first (which I think is the less complicated ordering), I'll help with fixing those!

[ahbarnett]

Got it :) Libin can also help. Sorry, returning to your review after a big interruption yesterday...

…

On Tue, Oct 15, 2024 at 6:28 AM mreineck ***@***.***> wrote: This PR will almist certainly conflict with #567 <#567>. If #567 <#567> is merged first (which I think is the less complicated ordering), I'll help with fixing those! — Reply to this email directly, view it on GitHub <#541 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACNZRST3BMQPNF4AJ2EIWX3Z3TU43AVCNFSM6AAAAABNIB4UP2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMJTGUYDKNZYHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

-- *-------------------------------------------------------------------~^`^~._.~' |\ Alex Barnett Center for Computational Mathematics, Flatiron Institute | \ http://users.flatironinstitute.org/~ahb 646-876-5942

[DiamonDinoia]

I think now is a good time to merge this.

[ahbarnett]

Thanks, very nice.