<div dir="ltr">Yes. I had a discussion with Noemi about using it with PETSc. The big requirement is formulating the discretization<div>in FEniCS. I think it might be possible to peel back one layer of interface and use it directly. Once I find the right</div><div>student, I will have them investigate. There are some very nice pieces in there.</div><div><br></div><div> Thanks,</div><div><br></div><div> Matt</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Feb 5, 2020 at 7:00 PM Smith, Barry F. via petsc-dev <<a href="mailto:petsc-dev@mcs.anl.gov">petsc-dev@mcs.anl.gov</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div style="word-wrap:break-word">
<div><br>
</div>
Lois sent out this announcement on hIPPYlib 3.0
<div><br>
<blockquote type="cite">
<div>Begin forwarded message:</div>
<br>
<div style="margin:0px">
<span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif;color:rgb(0,0,0)"><b>From:
</b></span><span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif">"McInnes, Lois Curfman" <<a href="mailto:curfman@anl.gov" target="_blank">curfman@anl.gov</a>><br>
</span></div>
<div style="margin:0px">
<span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif;color:rgb(0,0,0)"><b>Subject:
</b></span><span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif"><b>FW: [SIAM-CSE] Introducing hIPPYlib, a python-based inverse problems solver library</b><br>
</span></div>
<div style="margin:0px">
<span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif;color:rgb(0,0,0)"><b>Date:
</b></span><span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif">February 4, 2020 at 8:52:46 AM CST<br>
</span></div>
<div style="margin:0px">
<span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif;color:rgb(0,0,0)"><b>To:
</b></span><span style="font-family:-webkit-system-font,"Helvetica Neue",Helvetica,sans-serif">"Smith, Barry F." <<a href="mailto:bsmith@mcs.anl.gov" target="_blank">bsmith@mcs.anl.gov</a>><br>
</span></div>
<br>
<div>
<div>Have you seen this?<br>
<br>
On 2/4/20, 9:49 AM, "SIAM-CSE on behalf of Noemi Petra" <<a href="mailto:siam-cse-bounces@siam.org" target="_blank">siam-cse-bounces@siam.org</a> on behalf of
<a href="mailto:npetra@ucmerced.edu" target="_blank">npetra@ucmerced.edu</a>> wrote:<br>
<br>
We are pleased to announce the availability of hIPPYlib, an extensible<br>
software framework for solving large-scale deterministic and Bayesian<br>
inverse problems governed by partial differential equations (PDEs)<br>
with (possibly) infinite-dimensional parameter fields. The development<br>
of this project is being supported by the National Science Foundation.<br>
<br>
The current version of hIPPYlib is 3.0 and can be downloaded from:<br>
<br>
<a href="https://hippylib.github.io" target="_blank">https://hippylib.github.io</a><br>
<br>
This computational tool implements state-of-the-art scalable<br>
adjoint-based algorithms for PDE-based deterministic and Bayesian<br>
inverse problems. It builds on FEniCS for the discretization of the<br>
PDE and on PETSc for scalable and efficient linear algebra operations<br>
and solvers.<br>
<br>
A few features worth highlighting include:<br>
<br>
- Friendly, compact, near-mathematical FEniCS notation to express,<br>
differentiate, and discretize the PDE forward model and likelihood<br>
function<br>
<br>
- Large-scale optimization algorithms, such as globalized inexact<br>
Newton-CG method, to solve the inverse problem<br>
<br>
- Randomized algorithms for trace estimation, eigenvalues and singular<br>
values decomposition<br>
<br>
- Scalable sampling of Gaussian random fields<br>
<br>
- Linearized Bayesian inversion with low-rank based representation of<br>
the posterior covariance<br>
<br>
- Hessian-informed MCMC algorithms to explore the posterior<br>
distribution<br>
<br>
- Forward propagation of uncertainty capabilities using Monte Carlo<br>
and Taylor expansion control variates<br>
<br>
For more details, please check out the manuscript:<br>
<br>
<a href="http://arxiv.org/abs/1909.03948" target="_blank">http://arxiv.org/abs/1909.03948</a><br>
<br>
For additional resources and tutorials please check out the teaching<br>
material from the 2018 Gene Golub SIAM Summer School on ``Inverse<br>
Problems: Systematic Integration of Data with Models under<br>
Uncertainty" available at <a href="http://g2s3.com" target="_blank">http://g2s3.com</a>.<br>
<br>
Umberto Villa, Noemi Petra and Omar Ghattas<br>
<br>
-- <br>
Noemi Petra, PhD<br>
<br>
Assistant Professor of Applied Mathematics<br>
SIAM Student Chapter Faculty Advisor<br>
University of California, Merced<br>
<a href="http://faculty.ucmerced.edu/npetra/" target="_blank">http://faculty.ucmerced.edu/npetra/</a><br>
<br>
_______________________________________________<br>
SIAM-CSE mailing list<br>
To post messages to the list please send them to: <a href="mailto:SIAM-CSE@siam.org" target="_blank">
SIAM-CSE@siam.org</a><br>
<a href="http://lists.siam.org/mailman/listinfo/siam-cse" target="_blank">http://lists.siam.org/mailman/listinfo/siam-cse</a><br>
<br>
<br>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div><div><br></div><div><a href="http://www.cse.buffalo.edu/~knepley/" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br></div></div></div></div></div></div></div>