# Computational Techniques

**2020 - 2021 series**

Wednesdays 1100-1200 via MS Teams Computational Techniques Team

Wednesdays 1100-1200 via MS Teams Computational Techniques Team

Attendance at these classes is compulsory for MSc students.

**First-term week 2 - 14th October 2020**

*Introduction to the Scientific Computing Research Technology Platform part 1* - Professor David Quigley

**First-term week 3 - 21st October 2020**

*Introduction to the Scientific Computing Research Technology Platform part 2* - Professor David Quigley

**First-term week 4 - 28th October 2020**

*Introduction to software development * - Dr Chris Brady and Dr Heather Ratcliffe

This talk based on a summary of these slides

**First-term week 5 - 4th November 2020**

* Version control and software sustainability* - Dr Chris Brady and Dr Heather Ratcliffe

**First-term week 6 - 11th November 2020**

*Introduction to bash scripting -Dr Paul Brown*

**First-term week 7 - 18th November 2020**

*Making webpages with interactive content - *Dr Paul Brown

**First-term week 8 - 25th November 2020**

*No seminar*

**First-term week 9 - 2nd December 2020**

* Using GPUs at Warwick* - Professor David Quigley

The notebook used to deliver this talk, as well as much more detailed notebooks on Python GPU programming are at github.com/WarwickRSE/gpuschool2018

**First-term week 10 - 9th December 2020**

*Machine learning with Julia* - Dr Sebastian Vollmer

**2019-2020 series**

Wednesdays 11am - 12 noon in D1.07

Wednesdays 11am - 12 noon in D1.07

Attendance at these classes is compulsory for MSc students.

**First-term week 2 - 9th October 2019**

*Introduction to the Scientific Computing Research Technology Platform part 1* - Professor David Quigley

**First-term week 3 - 16th October 2019**

*Introduction to the Scientific Computing Research Technology Platform part 2* - Professor David Quigley

**First-term week 4 - 23rd October 2019**

*Introduction to software development * - Dr Chris Brady and Dr Heather Ratcliffe

This talk based on a summary of these slides

**First-term week 5 - 30th October 2019**

* Version control and software sustainability* - Dr Chris Brady and Dr Heather Ratcliffe

**First-term week 6 - 6th November 2019**

*Introduction to bash scripting -Dr Paul Brown*

**First-term week 7 - 13th November 2019**

*Making webpages with interactive content - *Dr Paul Brown

**First-term week 8 - 20th November 2019**

*No seminar
*

**First-term week 9 - 27th November 2019**

* Using GPUs at Warwick* - Professor David Quigley

The notebook used to deliver this talk, as well as much more detailed notebooks on Python GPU programming are at github.com/WarwickRSE/gpuschool2018

**2018-2019 series**

**First-term week 2 - 10th October 2018**

*Introduction to the Scientific Computing Research Technology Platform*

Dr David Quigley

The University of Warwick runs eight Research Technology Platforms (RTPs) which serve the university research community by providing large-scale shared facilities across multiple academic departments. I will introduce the facilities available via the Scientific Computing RTP, including our managed Linux desktop environment and high-performance computing (HPC) systems. This will include mechanisms for gaining access, working effectively within a managed multi-user Linux environment and how to access support and training.

**First-term week 3 - 17th October 2018**

*Statistical Inference from genomic data*

Dr Jere Koskela

A technological revolution in genetics is underway. Genome sequencing is getting cheaper and cheaper, and will eventually become a routine part of healthcare. It has been predicted that within 15 years one billion human genomes will have been sequenced. This data is exciting because patterns of variation in DNA sequences between individuals contain information on a number of biological and demographic processes, such as mutation, natural selection, population sizes, and migration events. However, such vast quantities of data raise a number of statistical and computational challenges. I will discuss some of the statistical techniques that have been applied to address these problems, with a focus on Monte Carlo methods such as importance sampling, and Markov Chain Monte Carlo (MCMC). I will also introduce some of the population genetics models that are used to study the evolution of large, random-mating populations, making particular use of an area known as coalescent theory.

**First-term week 4 - 24th October 2018**

*Data driven modelling - a model of chromosome oscillations from data to bifurcations*

Professor Nigel Burroughs

Cell biology is often stated as the new 'physics', a rich field where physical theories can be developed to explain/predict biological processes, analogous to the quantum physics and relativity successes early last century. Realising this ambitious aim is however proving difficult, particularly since biological processes are out of equilibrium, are often highly stochastic involving small numbers of molecules and are highly complex, displaying a range of phenomenal self-organising dynamics. In this talk I will examine what it means to 'explain' biological processes, and how a range of physical techniques from stochastic simulations, dynamical systems analysis and computational statistics can be coupled together to address these complex questions. Examples will be drawn from cytoskeletal mechanics and cell division.

**First-term week 6 - 7th November 2018**

*Understanding human behaviour with data science*

Professor Tobias Preis

In this lecture, we will outline some recent highlights of our research, addressing two questions. Firstly, can big data resources provide insights into crises in financial markets? By analysing Google query volumes for search terms related to finance and views of Wikipedia articles, we find patterns which may be interpreted as early warning signs of stock market moves. Secondly, can we provide insight into international differences in economic wellbeing by comparing patterns of interaction with the Internet? To answer this question, we introduce a future-orientation index to quantify the degree to which Internet users seek more information about years in the future than years in the past. We analyse Google logs and find a striking correlation between the country's GDP and the predisposition of its inhabitants to look forward. Our results illustrate the potential that combining extensive behavioural data sets offers for a better understanding of large scale human economic behaviour.

**First-term week 7 - 14th November 2018**

*Statistical inference using Markov chain Monte Carlo*

Dr Jake Carson

In this lecture I will introduce some techniques for model fitting within a Bayesian framework and illustrate them with some simple examples. In particular I will focus on Markov chain Monte Carlo and related methods. I will attempt to explain how it works, why it is so commonly used and give some practical guidance on its implementation.

**First-term week 8 - 21st November 2018**

*Image-based modelling of cell dynamics*

Dr Sharon Collier

Modern live-cell fluorescence microscopy enables us to visualise dynamic cellular processes in unprecedented detail. I will present ongoing research projects which are concerned with bringing together i) image analysis methods to track cells and their movements, and quantify spatio-temporal patterns of fluorescently labelled cellular constituents, and ii) mathematical models to investigate regulatory mechanisms of cellular biochemistry and mechanics.

**First-term week 9 - 28th November 2018**

*Software Development for Academics*

Dr Heather Ratcliffe

Software development is much more than writing code - it's about producing professional, maintainable, understandable software. Planning and style, documentation, version control, packaging, licensing and more. This talk aims to introduce some of the vital tools you should know about as researchers-who-write-code, so that your're well equipped to read and learn further and use them. We'll discuss version control in the form of Git, a little about software licensing, some nice packages for documenting code, and briefly mention a few other things that everybody should know, but nobody thinks to mention.

**2017-2018 series**

**First-term week 2 - 11th October 2017**

*Tackling complexity and self-organisation in biological systems*

Professor Nigel Burroughs

Cell biology is often stated as the new 'physics', a rich field where physical theories can be developed to explain/predict biological processes, analogous to the quantum physics and relativity successes early last century. Realising this ambitious aim is however proving difficult, particularly since biological processes are out of equilibrium, are often highly stochastic involving small numbers of molecules and are highly complex, displaying a range of phenomenal self-organising dynamics. In this talk I will examine what it means to 'explain' biological processes, and how a range of physical techniques from stochastic simulations, dynamical systems analysis and computational statistics can be coupled together to address these complex questions. Examples will be drawn from cytoskeletal mechanics and cell division.

**First-term week 3 - 18th October 2017**

*Statistical inference using Markov chain Monte Carlo*

Dr Simon Spencer

In this lecture I will introduce some techniques for model fitting within a Bayesian framework and illustrate them with some simple examples. In particular I will focus on Markov chain Monte Carlo and related methods. I will attempt to explain how it works, why it is so commonly used and give some practical guidance on its implementation.

**First-term week 4 - 25th October 2017**

*Understanding human behaviour with data science*

Professor Tobias Preis

In this lecture, we will outline some recent highlights of our research, addressing two questions. Firstly, can big data resources provide insights into crises in financial markets? By analysing Google query volumes for search terms related to finance and views of Wikipedia articles, we find patterns which may be interpreted as early warning signs of stock market moves. Secondly, can we provide insight into international differences in economic wellbeing by comparing patterns of interaction with the Internet? To answer this question, we introduce a future-orientation index to quantify the degree to which Internet users seek more information about years in the future than years in the past. We analyse Google logs and find a striking correlation between the country's GDP and the predisposition of its inhabitants to look forward. Our results illustrate the potential that combining extensive behavioural data sets offers for a better understanding of large scale human economic behaviour.

**First-term week 6 - 8th November 2017**

*Scientific Computing with Julia*

Professor Christoph Ortner

PART 1: Introduction to Julia. I will briefly introduce the language Julia and some of its tools, and show how it interpolates Matlab, Python and Lisp into a programming environment that is perfectly suited for numerically intensive computing, both rapid prototyping and HPC.

PART 2: I will show some examples from my own research on multi-scale materials modelling.

**First-term week 7 - 15th November 2017
**

*Stochastic Simulations*Professor Matthew Keeling

In this lecture we will initially discuss the importance of stochasticity in understanding real world problems. Stochasticity can be incorporated in many ways, but we will focus on individual-based, event-drive stochasticity and will discuss methods of simulating such dynamics. Unashamedly taking examples exclusively from ecology and epidemiology, we will consider both Gillespie’s Methods and Ensemble/Master equations. We will discuss what happens when population sizes become large — and approximations that make the problem computationally tractable. Finally we’ll look at fully individual-based spatial simulations and discuss methods that can provide a huge computational saving.

**First-term week 8 - 22nd November 2017**

*Image-based modelling of cell dynamics *

Professor Till Bretschneider

Modern live-cell fluorescence microscopy enables us to visualise dynamic cellular processes in unprecedented detail. I will present ongoing research projects which are concerned with bringing together i) image analysis methods to track cells and their movements, and quantify spatio-temporal patterns of fluorescently labelled cellular constituents, and ii) mathematical models to investigate regulatory mechanisms of cellular biochemistry and mechanics.

**First-term week 9 - 29th November 2017**

*Statistical Inference from genomic data*

Dr Paul Jenkins

A technological revolution in genetics is underway. Genome sequencing is getting cheaper and cheaper, and will eventually become a routine part of healthcare. It has been predicted that within 15 years one billion human genomes will have been sequenced. This data is exciting because patterns of variation in DNA sequences between individuals contain information on a number of biological and demographic processes, such as mutation, natural selection, population sizes, and migration events. However, such vast quantities of data raise a number of statistical and computational challenges. I will discuss some of the statistical techniques that have been applied to address these problems, with a focus on Monte Carlo methods such as importance sampling, Markov Chain Monte Carlo (MCMC), and Approximate Bayesian Computation (ABC). I will also introduce some of the population genetics models that are used to study the evolution of large, random-mating populations, making particular use of an area known as coalescent theory.

**2016-2017 series**

This course takes place on Wednesdays 11am-12noon in the D1.07 room.

** First-term week 2 - 12th October 2016 **

* Understanding human behaviour with data science *

Dr Tobias Preis

In this lecture, we will outline some recent highlights of our research, addressing two questions. Firstly, can big data resources provide insights into crises in financial markets? By analysing Google query volumes for search terms related to finance and views of Wikipedia articles, we find patterns which may be interpreted as early warning signs of stock market moves. Secondly, can we provide insight into international differences in economic wellbeing by comparing patterns of interaction with the Internet? To answer this question, we introduce a future-orientation index to quantify the degree to which Internet users seek more information about years in the future than years in the past. We analyse Google logs and find a striking correlation between the country's GDP and the predisposition of its inhabitants to look forward. Our results illustrate the potential that combining extensive behavioural data sets offers for a better understanding of large scale human economic behaviour.

** First-term week 3 - 19th October 2016 **

* Stochastic Simulations *

Professor Matthew Keeling

In this lecture we will initially discuss the importance of stochasticity in understanding real world problems. Stochasticity can be incorporated in many ways, but we will focus on individual-based, event-drive stochasticity and will discuss methods of simulating such dynamics. Unashamedly taking examples exclusively from ecology and epidemiology, we will consider both Gillespie’s Methods and Ensemble/Master equations. We will discuss what happens when population sizes become large — and approximations that make the problem computationally tractable. Finally we’ll look at fully individual-based spatial simulations and discuss methods that can provide a huge computational saving.

** First-term week 4 - 26th October 2016 **

* Big data and bioinformatics*

Dr Richard Savage

Medicine and biology are undergoing a data revolution. From whole-genome sequencing to digital imaging and electronic health records, new sources of data are promising to revolutionise how we treat disease and conduct our biomedical research. With these opportunities, however, come significant challenges. The data are often high-dimensional, noisy, with complex underlying structure. And we may wish to combine multiple data types from very different sources. I'll give a tour of some of this issues, focusing on some real-world projects that have the potential to change the way we do research in these areas. I'll also talk about how this relates to Warwick's involvement in large scale projects such as the 100,000 Genomes Project and the Alan Turing Institute.

** First-term week 5 - 2nd November 2016 **

* Computational techniques in mathematical biology *

Dr Nabil-Fareed Alikhan, Dr Till Bretschneider and Dr Giorgos Minas

** First-term week 6 - 9th November 2016 **

* No seminar this week *

** First-term week 7 - 16th November 2016 **

* Statistical inference using Markov chain Monte Carlo *

Dr Simon Spencer

In this lecture I will introduce some techniques for model fitting within a Bayesian framework and illustrate them with some simple examples. In particular I will focus on Markov chain Monte Carlo and related methods. I will attempt to explain how it works, why it is so commonly used and give some practical guidance on its implementation.

** First-term week 8 - 23rd November 2016 **

* Scientific Computing with Julia *

Professor Christoph Ortner

PART 1: Introduction to Julia. I will briefly introduce the language Julia and some of its tools, and show how it interpolates Matlab, Python and Lisp into a programming environment that is perfectly suited for numerically intensive computing, both rapid prototyping and HPC.

PART 2: I will show some examples from my own research on multi-scale materials modelling.

** First-term week 9 - 30th November 2016 **

* Inference and fitting of spatial dynamic systems in cell biology. *

Professor Nigel Burroughs

Cell biology is often stated as the new 'physics', a rich field where physical theories can be developed to explain/predict biological processes, analogous to the quantum physics and relativity successes early last century. Realising this ambitious aim is however proving difficult, particularly since biological processes are out of equilibrium, are often highly stochastic involving small numbers of molecules and are highly complex, displaying a range of phenomenal self-organising dynamics. In this talk I will examine what it means to 'explain' biological processes, including discussion of the types of models/modelling and when they are useful, comparing those models to data (reverse engineering) and verification of those models. Examples will will be drawn from cytoskeletal processes and cell division.

**2015-2016 series**

This course takes place on Wednesdays 11am-12noon in the Complexity lecture room.

** First-term week 2 - 14th October 2015 **

* Stochastic simulations *

Professor Matthew Keeling

In this lecture we will initially discuss the importance of stochasticity in understanding real world problems. Stochasticity can be incorporated in many ways, but we will focus on individual-based, event-drive stochasticity and will discuss methods of simulating such dynamics. Unashamedly taking examples exclusively from ecology and epidemiology, we will consider both Gillespie’s Methods and Ensemble/Master equations. We will discuss what happens when population sizes become large — and approximations that make the problem computationally tractable. Finally we’ll look at fully individual-based spatial simulations and discuss methods that can provide a huge computational saving.

** First-term week 3 - 21st October 2015 **

* Scientific Computing at Warwick: a double perspective from a Chemist/Director *

Professor Mark Rodger

This talk will take a look at Scientific Computing at Warwick from both a general and a personal perspective. From the general perspective I will seek to give an overview of the range of activities that go on within Warwick, of the role of the Centre for Scientific Computing in fostering those activities, and of some of the hardware and software that is readily available to assist research in the general area of Scientific Computing. To provide a more personal perspective, I will go on to describe some of the research that I do in the area of classical statistical mechanics and molecular modelling, in particular describing some of the adaptive molecular dynamics methods that have been developed in recent years to improve phase space exploration, characterisation of free energy landscapes, and the simulation of rare events for applications to materials science.

** First-term week 4 - 28th October 2015 **

* Inference and fitting of spatial dynamic systems in cell biology. *

Professor Nigel Burroughs

Cell biology is often stated as the new 'physics', a rich field where physical theories can be developed to explain/predict biological processes, analogous to the quantum physics and relativity successes early last century. Realising this ambitious aim is however proving difficult, particularly since biological processes are out of equilibrium, are often highly stochastic involving small numbers of molecules and are highly complex, displaying a range of phenomenal self-organising dynamics. In this talk I will examine what it means to 'explain' biological processes, including discussion of the types of models/modelling and when they are useful, comparing those models to data (reverse engineering) and verification of those models. Examples will will be drawn from cytoskeletal processes and cell division.

** First-term week 5 - 4th November 2015 **

* Mass Univariate and Multivariate Approaches to Understanding Genetic Variation in the Brain *

Professor Thomas Nichols

There has been great interest in discovering and understanding the role of genetic variation in brain imaging data. Typical "imaging genetics" studies use a small number of candidate genes, a small number of brain regions, or both. In this talk I will consider methods for searching for gene-brain associations over the entire genome and all brain regions. Such an approach presents massive computational and statistical challenges. I'll discuss two approaches, a mass-univariate approach and a multivariate approach. A mass-univariate model is the standard tool in neuroimaging analysis, but scaling it up for 100,000 SNPs requires a series of computational and statistical innovations. With our method applied to Tensor-Based Morphometry data from the ADNI project, we report the first gene-brain association to survive whole-genome, whole-brain familywise error correction. Our multivariate approach uses a Sparse Reduced Rank Regression (sRRR) to jointly and parsimoniously explain gene-brain associations. Detailed detailed power analyses show that the multivariate approach should have even greater power than the univariate approach.

** First-term week 6 - 11th November 2015 **

* Statistical inference using Markov chain Monte Carlo *

Dr Simon Spencer

In this lecture I will introduce some techniques for model fitting within a Bayesian framework and illustrate them with some simple examples. In particular I will focus on Markov chain Monte Carlo and related methods. I will attempt to explain how it works, why it is so commonly used and give some practical guidance on its implementation.

** First-term week 7 - 18th November 2015 **

* Big data and bioinformatics *

Dr Richard Savage

Medicine and biology are undergoing a data revolution. From whole-genome sequencing to digital imaging and electronic health records, new sources of data are promising to revolutionise how we treat disease and conduct our biomedical research. With these opportunities, however, come significant challenges. The data are often high-dimensional, noisy, with complex underlying structure. And we may wish to combine multiple data types from very different sources. I'll give a tour of some of this issues, focusing on some real-world projects that have the potential to change the way we do research in these areas. I'll also talk about how this relates to Warwick's involvement in large scale projects such as the 100,000 Genomes Project and the Alan Turing Institute.

** First-term week 8 - 25th December 2015 **

* No talk this week*

** First-term week 9 - 2nd December 2015 **

* Understanding human behaviour with data science *

Dr Tobias Preis

In this lecture, we will outline some recent highlights of our research, addressing two questions. Firstly, can big data resources provide insights into crises in financial markets? By analysing Google query volumes for search terms related to finance and views of Wikipedia articles, we find patterns which may be interpreted as early warning signs of stock market moves. Secondly, can we provide insight into international differences in economic wellbeing by comparing patterns of interaction with the Internet? To answer this question, we introduce a future-orientation index to quantify the degree to which Internet users seek more information about years in the future than years in the past. We analyse Google logs and find a striking correlation between the country's GDP and the predisposition of its inhabitants to look forward. Our results illustrate the potential that combining extensive behavioural data sets offers for a better understanding of large scale human economic behaviour.