New 3 yr post-doc in our Systems Neuroscience Group

"Generative models of brain disorders"

Applications open now -> July 15th

Twitter

Application

Course High-Performance Computing with Python

VIDEO OF THE COURSE: HIGH-PERFORMANCE COMPUTING WITH PYTHON 2021
GitHub

The material and videos for this year will be here shortly after the course.

PhD and postdoc Open positions at Technical University of Munich in Germany


Professorship of Computational Neurosciences
Prof. Dr. Julijana Gjorgjieva

Credit clcoding

Modeling Neural Circuits Made Simple

An updated draft of undergrad-level Computational Neuroscience textbook. @mitpress will keep an open access version available online.

Github

Do register for JuliaCon. You’ll receive the details on how to join the discord channel where all the conversation will happen during the conference, and other updates. https://juliacon.org/2022/tickets/

Julia workshop LIVE
https://www.youtube.com/watch?v=uiQpwMQZBTA

https://github.com/storopoli/Julia-Workshop

Job Ad:
I am looking for a postdoc researcher with an initial appointment of 1.5 years, which may be extendable to November 2025. The work location is the State University of New York at Buffalo, Buffalo, NY, USA. The PI is Naoki Masuda

www.naokimasuda.net
https://lnkd.in/gtUrtnhi

and the postdoc will be housed in the Department of Mathematics, which I belong to.
The project is on the "energy landscape analysis" of neuroimaging (particularly functional MRI, also perhaps MEG) data.
We have been developing and applying this analysis tool for neuroimaging data
With this project, we will further deploy and develop (depending on the candidate's expertise) these methods to various data in health and disease contexts. The original method is founded on the Ising model and somewhat related to networks. Our plan is to not to stick to the Ising model and try to incorporate other techniques including temporal network methods, which is my expertise.

Experience in data analysis (including PhD in a related field such as network science, (computational) brain/neuro- science, medicine, biology, computer science, applied mathematics, physics) and good communication skills are indispensable. However, expertise in fMRI data analysis, neuroscience, statistical physics underlying the energy landscape analysis (undergraduate level or higher statistical physics knowledge), and/or applied mathematics is not mandatory, while it is certainly a plus.

If you are interested or have any queries, please send your CV to Naoki Masuda at naokimas@gmail.com. I prefer the job start date to be asap but am flexible.

Link to post

Making specific sound when code reaches a line.


>>> import chime

>>> chime.success()
>>> chime.warning()
>>> chime.error()
>>> chime.info()


Github

Installing Julia on Ubuntu, evoke from terminal:
https://t.me/scientific_programming

Julia for High-Performance Computing | JuliaCon 2022
LINK

Making a package in Julia using PkgTemplates.
t = Template(user="YOUR_GITHUB_USERNAME", author="YOUR NAME", dir="DIRECTORY TO BUILD THE PACKAGE", julia=VERSION,
plugins=[
License(; name="MIT"),
Git(),
GitHubActions(),
Documenter{GitHubActions}()
]
)
How to develop a Julia package

The 9th annual JuliaCon, 2022
Recorded talks.

Dynamical Systems

YouTube

- Introduction, Documentation and Installation
- Dynamical Systems Base
- Chaos Tools
- Delay Embeddings
- Entropies and Dimensions
- Recurrence Analysis

GitHub

Some of Machine learning books.

Credit: svpino

PySINDy is a sparse regression package with several implementations for the Sparse Identification of Nonlinear Dynamical systems (SINDy).

In brief:
You have a bunch of time series of a dynamical system. And have some guess that what kind of terms it could have, polynomial, trigonometric, etc ...
Just need to give the time series to the program. It search over all possible options to finally pick terms that could produce the input time series, the package uses different optimization algorithms to fit the model to the data and also some penalty terms to reduce the number of terms as much as possible.
The interesting point is that it can also handle noisy time series.
combine with other packages like sklearn and many more...

SINDy GitHub
Tutorial Videos: YouTube

Recently some new algorithms also have been developed SymINDy.

The map() function in #Python takes a function and a series of arguments, and makes an iterable of results. It can also work on functions with multiple arguments.
(But most Python people prefer list comprehensions.)

Credit: nedbat

Have you tried out the sh package for #Python yet? This package makes calling Linux and Mac terminal commands really easy! 🐍🔥

pypi.org/project/sh/

Credit: driscollis

Job ad:
a talented post-doctoral researcher to lead exciting new methods and software developments for naturalistic neuroscience
The position is available immediately and for up to 4 years.

here

New on the @arxiv: “Ordered community detection in directed networks”

https://arxiv.org/abs/2203.16460

Py code

A short thread... https://t.co/NoCwFbOZJE