owned this note   changed 9 months ago  
Published Linked with GitHub

CUQIpy software training, 16–17 Sept 2024

UQIPI24 workshop, ICMS, Bayes Centre, Edinburgh, UK

Tentative programme

Monday 16 September

Afternoon

Time Activity
13:00 Welcome and Introduction (Jakob)
13:15 Overview of CUQIpy, Installation (Amal)
13:50 Probably the simplest BIP in the world (Amal)
14:10 Coffee break
14:25 Distributions (Charlie)
15:05 Forward models (Charlie)
15:45 Coffee break
16:00 Bayesian inverse problems (Nicolai)
16:50 Wrap-up before dinner (Nicolai)
17:00 End of afternoon session

Evening

Time Activity
17:00 Pizza night at Bayes Centre
18:00 CUQIpy mini-hackathon intro (Jakob)
18:15 Participant ideas and group formation
18:45 - ? Work in hackathon groups

Tuesday 16 December

Morning

Time Activity
09:30 Recap and plan for the day (Jakob)
09:40 Advanced features overview (multiple speakers)
10:10 Work session
10:40 Coffee break
10:55 Work session
11:45 Wrap-up and evaluation (Jakob)
12:00 End of Day 2

Resources

Plugins

Running the training notebooks

We present three approaches to run the training notebooks

Installing CUQIpy and running the notebooks on your local machine

Approach #2: Quick start

Launching the notebooks in the cloud (Google colab and binder)

Approach #3: Technical University of Denmark (DTU) servers

Explained in the next seciton

DTU Jupyter notebook servers

If you are not installing CUQIpy on own computer, you can use one of the two learnmore servers (ask CUQIpy team at event for access):

When logged in to the learnmore servers you should see a "launcher" tab with some icons to start new notebooks with different python packages installed. You can use the following ones:

  • CUQIpy
  • CIL (has CUQIpy-CIL plugin)
  • CUQIpy-PyTorch

In the launcher you also see an icon "terminal", click that to start a linux terminal. In the you type/copy-paste:

git clone git@github.com:CUQI-DTU/CUQI-Book.git

This will clone/copy the CUQI book to your directory, then you can navigate in the left menu into the training folder and start the training notebooks. You need to change the "kernel" in the top right corner to e.g. the "CUQIpy-env" one, in order for the notebook to have access to the python packages installed in there such as CUQIpy.

Ideas for hackathon projects

List project topics here and please add your name to any project you are interested in working on:

1. Computed tomography with CUQIpy-CIL (proposed by Margaret Duff)
margaret

​​​​- Margaret Duff (CIL)
​​​​- Richard Huber
​​​​- Jonas Bresch
​​​​- Yiyao Zhang (Liverpool)
​​​​- ...

2. PDEs with CUQIpy-FEniCS

​​​​- name 1
​​​​- ...

3. PSF deconvolution on astronomical images (Proposed by Charlie MacMahon-Geller)
charlie_markus

​​​​- Charlie MacMahon-Geller (Newcastle University)
​​​​- Markus Rau (Newcastle University)
​​​​- ...

4. Gravity anomaly (proposed by Lara Baalbaki)
lara

​​​​- Lara Baalbaki (DTU)
​​​​- ...

5. Surface reconstruction (proposed by Yiyao Zhang)
20240912_SlidePitch_Yiyao_Compressed

​​​​- Yiyao Zhang (Liverpool)
​​​​- ...

6. Multistatic radar (Proposed by Joshua Hellier)
joshua_triplet_circling

​​​​- Joshua Hellier (Manchester)
​​​​- ...

7. Semi-blind image deconvolution (proposed by Monica Pragliola)
monica

​​​​- Monica Pragliola (Bologna)
​​​​- ...

8. Phase retrieval (proposed by Dawit Hailu)
dawit

​​​​- Dawit Hailu
​​​​- ...

9. [insert topic title here]

​​​​- add name
​​​​- ...

Tuesday morning: Advanced features overview

CUQIpy Benchmarks

Naoki Sakai and Tania Goia, DTU CUQI summer interns

Implicit priors with regularized Gaussian distributions

Jasper Everink

Moreau-Yoshida Unadjusted Langevin Algorithm (MYULA) and Plug & Play priors Unadjusted Langevin Algorithm (PnPULA)

Rémi Laumont

conda create -n cuqipy-pnpula pip
conda activate cuqipy-pnpula
git clone git@github.com:remilaumont/CUQIpy.git
cd CUQIpy
git checkout pnpula
pip install ./
pip install deepinv scikit-image
  • Then the notebooks can be found at demos/cuqi_myula.ipynb and demos/cuqi_pnpula.ipynb.

CT with CUQIpy-CIL

Jakob S. Jørgensen

CUQIpy-PyTorch

Jakob S. Jørgensen

UQ for PDEs

Amal M. Alghamdi

Gibbs sampling

Chao Zhang

User showcases

Chao Zhang

Feedback

  • We greatly appreciate it if you can take few moments to let us know about your CUQIpy training experience in this post-training feedback form: https://tinyurl.com/cuqipy24uqipi

Acknowledgments

  • This work was supported by The Villum Foundation (grant no. 25893).
  • Thanks to all CUQI project members for valuable input and contributions to CUQIpy!
  • And thank you for participating in the CUQIpy training! Please let us know what you think about the software and the training using the the feedback link just above.

Software issues, problems, comments:

Please let us know about any and all problems, suggestions, etc. about CUQIpy that you may have:

Did you discover any bugs or strange behaviour?

  • …

Feature wishlist

  • …

Documentation issues

  • …

Other

  • …
Select a repo