Peter Mann

Lecturer in Computer Science at the University of St Andrews. Higher-order networks.

[Maarten Mortier @maartengm]

Let's see where this goes @stevenstrogatz , I have two chatbots playing your Romeo and Julia differential equation. In my first set of tweets on this I had Julia's behaviour flipped and they ended in romantic bliss!

[Peter Mann @PMannStAndrews]

Our new paper exploring Belief propagation on random and empirical graphs with edge-disjoint cliques and chordless cycles is out! Phys. Rev. E 107, 054303 (2023) - https://t.co/iO04YHf2BJ

[Rich Pitts 🇺🇦🐈‍⬛🏴‍☠️🔬 @RichPitts]

Delighted to see Prof Simon Dobson and @PMannStAndrews from @StAndrewsUniv have presented at https://t.co/pJDcFp0pbt work from their @OracleResearch project - And first cloud runs of https://t.co/KQkFaV2NlR an #epidemic #Simulation looking forward to so much more 🥰

[Vermont Complex Systems Institute @ UVM @uvmcomplexity]

"Temporal and probabilistic forecasts of epidemic interventions"

New preprint from @uvmmathstats PhD student @MariahBoudreau w/@andreaJoanAllen @_NJ_Roberts @all_are @LHDnets

https://t.co/T093YaCnnf

[NetScience @net_science]

Fitting degree distributions of complex networks. (arXiv:2212.06649v1 [physics.soc-ph]) https://t.co/kjwHDKbsWb

[Peter Mann @PMannStAndrews]

"Can you write a Python function that uses this function to perform many repeats of this experiment, saving the results into a list and then return the average and the variance of the repeats?". Pretty impressive!

[Peter Mann @PMannStAndrews]

"Write a Python function that, given a graph and a bond occupancy probability, T, performs bond percolation using networkx and returns the size of the largest connected component" ... Wow!

[Peter Mann @PMannStAndrews]

And here is its output for a random graph

[James Gleeson @gleesonj]

Application portal for PhD positions is now open, please RT

[Leah Keating @leahakeating]

This work with @DavidJPOS and @gleesonj is now on arXiv (https://t.co/kpO9lqCdIp):

"A generating-function approach to modelling complex contagion on clustered networks with multi-type branching processes"

Here is a quick overview of our paper: (1/6)

[NetScience @net_science]

A generating-function approach to modelling complex contagion on clustered networks with multi-type branching processes. (arXiv:2212.00584v1 [physics.soc-ph]) https://t.co/Hxq17fwV6n

[Peter Mann @PMannStAndrews]

Is the symbol ‘§’ really important enough to occupy absolute prime Mac keyboard territory right next to number 1 and tab?? The number of accidental §-ing into my terminal is unreal.

[NetSci 2023 @netsci2023]

Hello World!

After 3 years, NetSci is finally back in person!

Vienna, July 10-14 2023.

https://t.co/F4fJ0lRyeS

The conference will be held in the historic main building of the University of Vienna.

#NetSci2023

@dnds_ceu @ceu @CSHVienna @netscisociety

[Complexity Digest @cxdig]

Statistical inference links data and theory in network science https://t.co/QK45xpfQzH

[PyG @PyG_Team]

GNNs have lots of use cases in the world of physics and simulation. Graphs can express a complex, dynamic system with many moving parts.
Great blog on how PyG can be used to tackle some of these problems: https://t.co/uHuz6TPW1d

[NetScience @net_science]

Spreading Processes with Mutations over Multi-layer Networks. (arXiv:2210.05051v1 [physics.soc-ph]) https://t.co/vOhZEbuvDq

[Peter Mann @PMannStAndrews]

Our new arXiv pre-print examines the role of different mixing patterns between vertices in higher-order networks. Applying our model to empirical networks, we found some pretty cool results - beyond dyadic interactions are really important! https://t.co/ytlK0T1IqU

[Peter Mann @PMannStAndrews]

@Azumanga Not really sure of specific guides, but I find poetry really good for handling environments and versioning; flake8 or black for pre-commit linting; GitHub actions for YAML based CI pipelines (test suite etc). For docs, sphinx. And an up to date makefile as I forget commands :D

[Peter Mann @PMannStAndrews]

@dorogovtsev Congratulations on your book! Your video worked, I look forward to reading it when it arrives :D

[Peter Mann @PMannStAndrews]

@EpiEllie I like using Newman-Ziff for bond percolation, under the assumption that bond percolation is equatable to Gillespie simulation or similar.