Research Computing Summer School 2018

The Research Computing Service, the Computational Methods Hub and Imperial Software Carpentry are happy to present the Research Computing Summer School 2018.  This event brings together the scientific community, external lecturers and the Research Computing team for four days of tutorials, lectures and exchange of ideas.

This year, we decided to celebrate the Imperial Software Carpentry initiative with an extended workshop.  Software Carpentry is an international organisation of certified volunteer instructors that teaches research computing.  The teaching materials are centrally developed and maintained and presented as two-day introductory workshops that follow a well-established organisational blueprint with a clear mission “to help scientists and engineers get more research done in less time and with less pain by teaching them basic lab skills for scientific computing”.  The College now has 14 certified instructors and we have been offering regular workshops since February 2018.

The summer school workshop will teach the basis of Bash shell, Git version system, SQL, database design, basis of R and Python.  In addition, we included an advanced day on Scientific Python and R.  

The summer school will conclude with an introduction to the College's new research data storage service and Imperial Research Software Engineering Showcase.  We are also pleased to welcome Alces who will describe the latest developments in high performance computing in the Cloud.

Stay tuned for more details!  If you have suggestions or would like to participate, please email

 

Dates: September 24-27, 2018
Venue: Huxley 340 and 341  (building 13 on the map), Sir Ernst Chain 311 (building 30 on the map)
Campus: Imperial College London, South Kensington campus
Organizers: Research Computing ServiceComputational Methods Hub and Imperial Software Carpentry
Registration: Registration link.  You can register for the whole week or for individual events separately.
Questions: Katerina Michalickova
Links:
Setup instructions, materials and collaborative documents
Additional workshop materials (internal)

Registration is now open! You can register for the whole week or for individual days (events) separately.

 

Programme:

Monday, September 24
Time Track 1, Huxley 340
PLEASE NOTE THE ROOM CHANGE
Track 2, Sir Ernst Chain 311
10:00-11:00 Bash Shell 
(file and directory management)
Amit Mandal, Heart and Lung Institute and Santiago Lacalle, Research Computing Service
THIS SESSION IS NOW FULLY BOOKED
-
11:00-13:00 Bash Shell 
(text editing, for loops, shell scripts)
Amit Mandal, Heart and Lung Institute and Santiago Lacalle, Research Computing Service
THIS SESSION IS NOW FULLY BOOKED
Introduction to SQL
Christopher Tomlinson, Surgery & Cancer
THIS SESSION IS NOW FULLY BOOKED
14:00-17:00 Git version control 
(managing a project with Git, collaborating using GitHub)
Lucy Whalley, Materials and Stefano Galvan, Mechanical Engineering
THIS SESSION IS NOW FULLY BOOKED
Database design 
(database server and client, relational database, tables, primary and foreign keys, managing one-to-one, one-to-many and many-to-many relationships)
Christopher Tomlinson, Surgery & Cancer
THIS SESSION IS NOW FULLY BOOKED
Tuesday, September 25
Time Track 1, Huxley 340 Track 2, Huxley 341
10:00-13:00 Programming with Python 1
(Jupyter Notebooks, data types, build-in functions, lists)
Katerina Michalickova, Research Computing Service
THIS SESSION IS NOW FULLY BOOKED
R for Reproducible Scientific Analysis 1
(R studio, variable types, getting help, lists, subsetting)
Anat Melamed, Medicine, Amit Mandal, Heart and Lung Institute and Barbara Bodinier, Epidemiology & Biostatistics
THIS SESSION IS NOW FULLY BOOKED
14:00-17:00 Programming with Python 2
(for loops, conditionals, functions, variable scope, introduction to Pandas)
Daniel Nolte, Bioengineering and Lucy Whalley, Materials
THIS SESSION IS NOW FULLY BOOKED
R for Reproducible Scientific Analysis 2
(control flow, creating functions, data frames)
Amit Mandal, Heart and Lung Institute and Barbara Bodinier, Epidemiology & Biostatistics
THIS SESSION IS NOW FULLY BOOKED
Wednesday, September 26
Time Track 1, Huxley 340 Track 2, Huxley 341
10:00-13:00 Scientific Python 1
(numpy, introduction to scipy, matplotlib)
Nicolas Barral, Earth Science and Engineering and Edward Smith, Mechanical Engineering
THIS SESSION IS NOW FULLY BOOKED
Data Science with R
(ggplot, tidyverse, dplyr)
Kiana West, Surgery & Cancer and Maryam Karimi, Epidemiology & Biostatistics
THIS SESSION IS NOW FULLY BOOKED
14:00-17:00 Scientific Python 2
(workflow for software quality assurance, building sustainable software)
Mayeul d'Avezac, Research Computing Service and Adam Jackson, UCL
THIS SESSION IS NOW FULLY BOOKED
Writing Reports with R
(knitr, R markdown, shiny)
Maryam Karimi, Epidemiology & Biostatistics and Anat Melamed, Medicine
THIS SESSION IS NOW FULLY BOOKED
Thursday, September 27
Time Huxley 340
10:00 - 12:00 High Performance Computing in the Cloud[info]
(when is it useful, where are we now, case studies, benchmarks)
Cristin Merritt, Wil Mayers, Alces Flight
13:00 -13:30 New Research Data Storage at the Reseach Computing Service 
Matthew Harvey, Research Computing Service
14:00 - 16:30 Imperial Research Software Engineering Showcase
chair: Jeremy Cohen, Computing
Programme
16:30 - 18:00 Reception

 

Workshops 

 

High Performance Computing in the Cloud

Exploring the adoption of cloud HPC through real case studies, open source, and free tools available on Amazon Web Services, Google Cloud Platform, and Microsoft Azure.

Cristin Merritt, Wil Mayers, Alces Flight

High Performance Computing (HPC) has a long and storied tradition born out of hardware.  With the introduction of cloud HPC over two years ago arguments broke out over how viable the technology would be and if there would be a mass abandonment of data centres and in-house facilities.  The result has been far more practical and inclusive than originally thought.  This sessions covers how cloud HPC is being adopted into the community and what tools can be used now on the major public cloud providers of Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure to get going in HPC.  Using real case studies we’ll plot out where cloud is working best for those wanting to embrace this new
technology.  Specific focus will be on open source, published templates and free subscription tools and applications to ensure those who attend can continue their exploration post-session.

Session break-down:

  • History of cloud HPC 2016 to present
    • Change management
    • Hardware v. Cloud
  • How cloud makes HPC more inclusive - four questions
    • What is your definition of cloud?
    • Do you have a pre-existing cloud vendor relationship?
    • Are you hardware v. cloud?
    • How much are you willing to let go of traditional HPC?
  • The technical progress of cloud HPC
    • The views of the cloud platforms on HPC, as interpreted by a partner
      • AWS
      • Google
      • Azure
    • HPC tools and services available
      • Open Source
      • Commercial
  • Alces Flight on cloud
    • Discussion of Flight on the cloud platforms - how to get going
      • Bypassing the upskilling
      • Templates - friend and foe
    • Demo of Flight Launch - which obfuscates the cloud platform and focuses more on the HPC aspects (also available for attendees to use for free during or post-session)