Reaching out to Francesco Faccio for help to test performance of Ode15{i,s} solvers
Sarrah Bastawala <[email protected]> Wed, 31 Aug 2022 18:51:56 +0530
| Newsgroups | gmane.comp.gnu.octave.maintainers |
|---|---|
| Message-ID | <CAEvqwU=zVPxxj6MoTrq5bs3CxNyO38UzhYNM7Zs6GPmeYNcs0g@mail.gmail.com> |
--===============4141587936437102132==
Content-Type: multipart/alternative; boundary="00000000000045db7505e789626a"
--00000000000045db7505e789626a
Content-Type: text/plain; charset="UTF-8"
Greetings Octave Developers!
This mail is primarily to reach out to Francesco Faccio, a 2016 GSoC
Student Developer
under GNU Octave, who worked on the Ode15{i,s} Solver Implementations
<https://gsoc2016ode15s.blogspot.com/2016/05/introducing-project.html> in
Octave.
I am pleased to inform that I have been continuing this implementation
under GSoC '22,
hoping to make the dependency of ode15{i,s} on KLU optional, by replacing
the N_Vector,
Sparse Jacobian and Solver by SUNDIALS with Octave's custom implementations
of the
same, under the mentorship of Carlo de Falco
<https://octave.discourse.group/u/cdf/summary>. The entire thread of the
project can be found here.
<http://https://octave.discourse.group/t/gsoc-2022-project-ode15-i-s/1343/143?u=sarrah-basta>
While working on these, we also hoped for some improvement in the
efficiency of the solvers,
as we would be avoiding making copies of data and also use our vectorized
operations in the
vector and matrices. Now that we have neared the end of the project, we
wish to test
the performance of the solver with these new implementations. Francesco had
made some
tests for this exact purpose, however the repository containing them listed
on the blog
seems to have been discontinued.
This is a humble request for him to reply if he has a copy of those
codes stored somewhere,
or access to the repository, as that would be a great help !
Awaiting a response in earnest,
Regards,
Sarrah Bastawala
Student, VJTI Mumbai
--00000000000045db7505e789626a
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
<div dir=3D"ltr">Greetings Octave Developers!<br><br>=C2=A0 =C2=A0 This mai=
l is primarily to reach out to Francesco Faccio, a 2016 GSoC Student Develo=
per<br>under GNU Octave, who worked on the=C2=A0<a href=3D"https://gsoc2016=
ode15s.blogspot.com/2016/05/introducing-project.html ">Ode15{i,s} Solver Im=
plementations</a>=C2=A0in Octave.<br><br>=C2=A0 =C2=A0 I am pleased to info=
rm that I have been continuing this implementation under GSoC '22,=C2=
=A0<br>hoping to make the dependency=C2=A0of ode15{i,s} on KLU optional, by=
replacing the N_Vector,=C2=A0<br>Sparse Jacobian and Solver by SUNDIALS wi=
th Octave's custom implementations of the<br>same, under the mentorship=
of <a href=3D"https://octave.discourse.group/u/cdf/summary">Carlo de Falco=
</a>. The entire thread of the project can be found=C2=A0<a href=3D"http:/=
/https://octave.discourse.group/t/gsoc-2022-project-ode15-i-s/1343/143?u=3D=
sarrah-basta">here.</a><br>=C2=A0 =C2=A0=C2=A0<div>=C2=A0 =C2=A0 =C2=A0Whil=
e working on these, we also hoped for some improvement in the efficiency of=
the solvers,<br>as we would be avoiding making copies of data and also use=
our vectorized operations in the=C2=A0<br>vector and matrices. Now that we=
have neared the end of the project, we wish to test=C2=A0<br>the performan=
ce of the solver with these new implementations. Francesco had made some=C2=
=A0<br>tests for this exact purpose, however the repository=C2=A0containing=
them listed on the blog=C2=A0<br>seems to have been discontinued.<br><br>=
=C2=A0 =C2=A0 This is a humble request for him to reply if he has a copy of=
those codes stored somewhere,<br>or access to the repository, as that woul=
d be a great help !<br><br>=C2=A0 =C2=A0 Awaiting a response in earnest,<br=
><br>Regards,<br>Sarrah Bastawala</div><div>Student, VJTI Mumbai=C2=A0</div=
></div>
--00000000000045db7505e789626a--
--===============4141587936437102132==
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: base64
Content-Disposition: inline
Ci0tLS0tLS0tLS0KV2UgYXJlIHRyYW5zaXRpb25pbmcgdG8gYSB3ZWIgYmFzZWQgZm9ydW0KZm9y
IG1haW50YWluZXIgZGlzY3Vzc2lvbnMgYXQKaHR0cHM6Ly9vY3RhdmUuZGlzY291cnNlLmdyb3Vw
L2MvbWFpbnRhaW5lcnMK
--===============4141587936437102132==--