- firedrake - imperial.ac.uk

vertical slice + topography
by Cotter, Colin J 11 Apr '14

11 Apr '14

Dear Firedrakers, The attached image doesn't look very exciting but it shows that we managed to get a hydrostatic pressure calculation in the presence of topography (a hill!) working in the extruded FEEC framework. Things are starting to work! We took a temperature field that depends only on z, and computed the pressure on a terrain following mesh, the exact solution depends only on z too. This is with the lowest order spaces i.e. DG1 for pressure and P0xP1 for temperature. cheers --cjc

2 1

Fwd: Tim Palmer, Grantham Seminar Speaker - 9 April
by David Ham 08 Apr '14

08 Apr '14

Hi All, An interesting talk about a highly unconventional approach to precision in algorithms tomorrow... David Greater accuracy with less precision: a new paradigm for weather and climate prediction. A link to the abstract and the bio is below: http://www3.imperial.ac.uk/newsandeventspggrp/imperialcollege/naturalscienc… We hope to see many of you here. Best wishes Gosia Gosia Gayer Grantham Institute for Climate Change Imperial College London South Kensington Campus London SW7 2AZ Tel: +44 (0)20 7594 9666 Fax: +44 (0)20 7594 9668 Email: g.gayer(a)imperial.ac.uk www.imperial.ac.uk/climatechange *From:* Gayer, Gosia *Sent:* 08 April 2014 14:13 *To:* Toumi, Ralf; Czaja, Arnaud (a.czaja(a)imperial.ac.uk) Voulgarakis, Apostolos (a.voulgarakis(a)imperial.ac.uk) Ham, David A ( david.ham(a)imperial.ac.uk) *Cc:* Nissan, Hannah; Hoskins, Brian J; Buckle, Simon J *Subject:* RE: Tim Palmer, Grantham Seminar Speaker - 9 April Apologies, the programme is now attached. Gosia *From:* Gayer, Gosia *Sent:* 08 April 2014 14:11 *To:* Toumi, Ralf; Czaja, Arnaud (a.czaja(a)imperial.ac.uk) Voulgarakis, Apostolos (a.voulgarakis(a)imperial.ac.uk) Ham, David A ( david.ham(a)imperial.ac.uk) *Cc:* Nissan, Hannah; Hoskins, Brian J; Buckle, Simon J *Subject:* RE: Tim Palmer, Grantham Seminar Speaker - 9 April Dear All Thank you for making yourselves available to meet Professor Tim Palmer. Please find attached Tim’s schedule for tomorrow. The meetings will be in the small office at the Grantham Institute. Best wishes Gosia ------------------------------ *From:* Gayer, Gosia *Sent:* 04 April 2014 11:19 *To:* Czaja, Arnaud; Brindley, Helen E; Graven, Heather; Toumi, Ralf; Piggott, Matthew D; Cotter, Colin J; Ham, David A *Subject:* Tim Palmer, Grantham Seminar Speaker - 9 April Dear All I am in the process of organising a programme for Professor Tim Palmer, who is our Grantham Seminar speaker on 9 April 2014. Title – “Greater accuracy with less precision: a new paradigm for weather and climate prediction”. *Speaker – **Professor Tim Palmer, Royal Society Research Professor in Climate Physics, University of Oxford* http://www3.imperial.ac.uk/newsandeventspggrp/imperialcollege/naturalscienc… We know that our speakers value the opportunity to talk to Imperial’s academics and I normally organise 3 x 30 minute sessions starting at 14.15 ie: 14.15-14.45 14.45-15.15 15.15-15.45 Your names were suggested by Professor Brian Hoskins or you may have expressed interest to meet Professor Palmer yourself. Pease let me know if you are willing and available to meet Professor Palmer. It would be very helpful to have your answer asap. Many thanks Gosia Gosia Gayer Grantham Institute for Climate Change Imperial College London South Kensington Campus London SW7 2AZ Tel: +44 (0)20 7594 9666 Fax: +44 (0)20 7594 9668 Email: g.gayer(a)imperial.ac.uk www.imperial.ac.uk/climatechange -- Dr David Ham Departments of Mathematics and Computing Imperial College London http://www.imperial.ac.uk/people/david.ham

1 0

New meeting time
by Lawrence Mitchell 03 Apr '14

03 Apr '14

Dear all, the firedrake status meeting has moved from Thursday at 3pm to Friday at 1pm in WPL 301. Hopefully this is still convenient for everyone. Lawrence

2 1

Re: [firedrake] R space
by David Ham 02 Apr '14

02 Apr '14

Hi All, Lawrence and I had a productive lunch in which we think we worked out how to do R efficiently. Conversation on this subject will now move to the issue (https://github.com/firedrakeproject/firedrake/issues/226) and all of your input is actively sought. David On 12 March 2014 09:21, Farrell, Patrick <patrick.farrell06(a)imperial.ac.uk>wrote: > On 12/03/14 09:17, Patrick Farrell wrote: > > R is useful for things other than boundary conditions, too. If you have > a constrained > > optimisation problem with a single constraint (an integral constraint, > say, like [0]), > > P.S. For working code where you need this, see > > > https://bitbucket.org/dolfin-adjoint/da-applications/src/3907be10e58ab0f61e… > > which solves the KKT equations associated with example 4 of [0] by means > of a combined continuation on log barrier terms and casting the problem > as a mixed complementarity problem to the PETSc VI solver. > > I'd be dead impressed if firedrake could do this (even more so if it could > do this on GPUs). > > Patrick > > _______________________________________________ > firedrake mailing list > firedrake(a)imperial.ac.uk > https://mailman.ic.ac.uk/mailman/listinfo/firedrake > -- Dr David Ham Departments of Mathematics and Computing Imperial College London http://www.imperial.ac.uk/people/david.ham

1 0

PADAL workshop on abstractions for locality
by Kelly, Paul H J 01 Apr '14

01 Apr '14

Dear OP2, PyOP2, Firedrake, AEcute and Pamela people I'm going to a workshop in Lugano at the end of the month, on programming abstractions for locality. I thought you might be interested in the topics of the other speakers: http://padalworkshop.org/ For example: Leonidas Linardakis, Memory Abstraction via Subsets in the ICON Domain Specific Language - works on the ICON weather model. Is this the guy we tried to set up an OP2 collaboration with? I have created an openly-shared Google Doc - I am inviting you to read through the abstracts and contribute any comments you may have: https://docs.google.com/document/d/19JotcaJ1S7DGZO91X90-yHUsPwagB2hBNg4Lsu8… Reply to this email when you comment so we are prompted to read what you wrote :) Many thanks Paul

1 0

environment
by Cotter, Colin J 27 Mar '14

27 Mar '14

Dear Firedrakers, I've opened a new terminal and there is something not right with my environment: cjc1@richardson:~/mercurial/vertical-tests$ python Traceback (most recent call last): File "/usr/lib/python2.7/site.py", line 68, in <module> import os File "/usr/lib/python2.7/os.py", line 49, in <module> import posixpath as path File "/usr/lib/python2.7/posixpath.py", line 17, in <module> import warnings File "/usr/lib/python2.7/warnings.py", line 8, in <module> import types File "/home/cjc1/firedrake/firedrake/firedrake/types.py", line 1, in <module> import numpy as np ImportError: No module named numpy I had this yesterday on my workstation but I've forgotten how I fixed it. What do I need to do? cheers --cjc

1 1

Fwd: K40
by Kelly, Paul H J 27 Mar '14

27 Mar '14

We now have access to a shiny new Nvidia K40 GPU on CX1, see below. This is a kind donation from Nvidia, as a result of Imperial becoming an official CUDA Research Centre (the process was led by Peter Vincent). Paul Begin forwarded message: From: "Harvey, M J" <m.j.harvey(a)imperial.ac.uk<mailto:m.j.harvey@imperial.ac.uk>> Date: 27 March 2014 16:40:53 GMT To: "Kelly, Paul H J" <p.kelly(a)imperial.ac.uk<mailto:p.kelly@imperial.ac.uk>>, "Vincent, Peter E" <p.vincent(a)imperial.ac.uk<mailto:p.vincent@imperial.ac.uk>> Subject: Re: K40 Paul, Yes. For the moment, you can get it with: #PBS -lselect=1:ncpus=1:ngpus=1:host=cx1-30-12-1 #PBS -q pqkelly ________________________________________ De: Kelly, Paul H J Enviado: jueves, 27 de marzo de 2014 16:36 Para: Vincent, Peter E CC: Harvey, M J Asunto: Re: K40 Hi Peter, hi Matt Did this card get installed? Cheers, Paul On 16 Mar 2014, at 23:46, Vincent, Peter E wrote: Dear All, Regarding below, the consensus was to put the card in a shared machine in HPC. I plan to meet Matt Harvey later this week to hand over the card. I will send out further details of how to access it once it is in place. Cheers Peter Dr Peter Vincent MSci ARCS DIC PhD Lecturer and EPSRC Early Career Fellow Department of Aeronautics Imperial College London South Kensington London SW7 2AZ UK web: www.imperial.ac.uk/aeronautics/research<http://www.imperial.ac.uk/aeronautics/research>/vincentlab<http://www.imperial.ac.uk/aeronautics/research/vincentlab> twitter: @Vincent_Lab<https://twitter.com/Vincent_Lab> On 10 Mar 2014, at 19:09, Peter Vincent <p.vincent(a)imperial.ac.uk<mailto:p.vincent@imperial.ac.uk>> wrote: Dear All, Nvidia have sent us a K40 as part of the CRC. Can I please have suggestions for what to do with it. Options include: 1.) Put it in a box that we can all access remotely. 2.) Pass it around between us so we can each play with it for a few weeks in our local systems. Cheers Peter Dr Peter Vincent MSci ARCS DIC PhD Lecturer and EPSRC Early Career Fellow Department of Aeronautics Imperial College London South Kensington London SW7 2AZ email: p.vincent(a)imperial.ac.uk<mailto:p.vincent@imperial.ac.uk> web: www.imperial.ac.uk/aeronautics/research/vincentlab<http://www.imperial.ac.uk/aeronautics/research/vincentlab>/<http://www.imperial.ac.uk/aeronautics/research/vincentlab/> twitter: @Vincent_Lab<https://twitter.com/#!/Vincent_Lab>

1 0

Expressions compiler
by Cotter, Colin J 27 Mar '14

27 Mar '14

Dear Firedrakers, Please excuse my excitement, but the expressions compiler is extremely cool. For example, it allows an application of an SSPRK timestepping scheme in the following compact form: while (t <= end): solve(lhsA == action(rhsA,oldT), newT) T1.assign(newT) solve(lhsA == action(rhsA,T1), newT) T2.assign(0.75*oldT + 0.25*newT) solve(lhsA == action(rhsA,T2), newT) newT.assign((1.0/3.0)*oldT + (2.0/3.0)*newT) oldT.assign(newT) cheers --cjc

1 0

Fwd: PDESoft Abstract
by Lawrence Mitchell 27 Mar '14

27 Mar '14

Begin forwarded message: > From: "Bercea, Gheorghe-teodor" <gheorghe-teodor.bercea08(a)imperial.ac.uk> > Subject: PDESoft Abstract > Date: 20 March 2014 16:28:51 GMT > To: "Ham, David A" <david.ham(a)imperial.ac.uk> > Cc: "Kelly, Paul H J" <p.kelly(a)imperial.ac.uk>, "McRae, Andrew" <a.mcrae12(a)imperial.ac.uk>, "Luporini, Fabio" <f.luporini12(a)imperial.ac.uk>, "Loriant, Nicolas" <n.loriant(a)imperial.ac.uk>, Lawrence Mitchell <lawrence.mitchell(a)ed.ac.uk> > > Hi everyone, > > Here's my abstract for PDESoft. It's a tentative abstract, it's longer than it should be, and the results section contains a bit of wishful thinking. With that in mind, I would like to know if the message matches the audience. > > Let me know what you think. > > Best wishes, > > Doru > >

1 0

FW: PRISM talk: Thursday 3rd April 12:00 RH254
by Cotter, Colin J 27 Mar '14

27 Mar '14

I'm away next week, but would be good if some others could get to this to help cement our college computational science connections. cheers --cjc ________________________________________ From: Chris Cantwell [c.cantwell(a)imperial.ac.uk] Sent: 27 March 2014 08:34 To: Sherwin, Spencer J; Kelly, Paul H J; Pain, Christopher; Piggott, Matthew D; Cotter, Colin J; Ham, David A; Peiro, Joaquim; Gorman, Gerard J; Vincent, Peter E Subject: PRISM talk: Thursday 3rd April 12:00 RH254 Dear all, We have a visitor, Kilian Lackhove from the Technische Universitat Darmstadt, funded under the PRISM platform visiting on Thursday and Friday next week and he will be giving a brief talk on his work. Title: "Using Nektar++ as CAA solver in a coupled setup for combustion noise modeling" Time: Thursday 1st April, 12:00 Room: Aeronautics, RH254 I would be grateful if you could circulate to your groups and anyone interested is welcome to attend. Please let me know if you would also like to meet with him during his visit. Cheers, Chris -- Chris Cantwell Research Fellow National Heart & Lung Institute Imperial College London South Kensington Campus Room L57, Roderic Hill Building, Department of Aeronautics London SW7 2AZ Email: c.cantwell(a)imperial.ac.uk www.imperial.ac.uk/people/c.cantwell

1 0