 I'm Dr Shamian Bennett. I'm an epidemiologist at the Australian National University and I've had a research team that's looking at the impacts of climate change on human health. The Australian National Data Service Ground has allowed us to develop a scientific workflow system to access and merge large datasets on our environment, population and health. And we've built our workflows using Keplo. We've called our workflow system SWISH, which stands for Scientific Workflow and Integration Software for Health, and it's freely available to other researchers from our website. To assess the impacts of climate change on health, you need to gather a large amount of data, including weather, definitions of extreme weather events, population data, mortality data including the cause of death, and all within a consistent spatial framework. And all of these steps have to happen before you actually get to the analysis of the health impacts linked to climate change. You can download complete workflows from our website that will set up your connection to the extreme weather events database. In this example, the actor GetWeather SLA connects to the database to locate daily weather data in the statistical local area of your choice. Double click on the string constant box and enter the name of the SLA that you want. In this example, we're looking at Kayleen. Then click commit. The second string constant shows you where the resulting file will be saved, in this case on the C drive in the temp folder. When you're done, click the run button in the toolbar and you'll see the word executing appear in the bottom left corner. When the results window appears, scroll down and check that a CSV file with the suburb name has been created. This workflow takes the data we obtained in the previous example and calculates how many days had an average temperature of above 35 degrees. You can use the search function to locate specific actors. To complete this workflow, it needs one more input using the string constant actor. You then drag and drop the actor into your workflow space and double click to edit the parameters. In this case, our value is 35 degrees. Then connect this box to the greater than actor using the connector arrows. To explore the many swish actors, go to the components window and open the folder called my workflows. Click the small plus sign to find the green actor icon, then drag and drop into your workflow. Here, we are adding an actor to the workflow to display our result at the end. When all of the actors are connected, click on the run button in the toolbar. The results are produced in the CSV file, which opens here in a text box, but can easily be opened in other statistical programs like Stata and R. Accessing data often requires high level programming expertise. Using swish can help reduce those barriers to data access by providing a user friendly drag and drop interface. Swish workflows are a single file, so it's very easy to share with other researchers. This can help us with collaborative research and improve the reproducibility of results. In addition, swish also packages multiple steps using Stata and R into a single executable flowchart that documents your working as you go. I'm Keith Theer. I'm a Biostatistician and Epidemiologist at the Australian National University. In 2007-2008, the Gano Climate Change Review Team asked a group of us here at NSF, the National Centre for Epidemiology and Population Health, to look at how health impacts of climate change might pan out in Australia out of the 2100. To do this, I used several kinds of data from different sources. The demographic breakdown of the Australian population by age, sex and location around the country, daily records of mortality which are collected within each state, and the daily weather records collected by the Bureau of Meteorology. Merging these various sources of data into a final dataset that I could use for a statistical analysis was complicated. I did this initially back in 2007 using scripts in the statistical package Stata which I wrote specifically for the purpose. The final script in Stata that did this operation was quite long, several hundred lines of Stata code. And it took me a lot of time to debug it, make sure that I was reasonably confident it was correct. And even when I was confident myself, it was difficult to communicate it to other potential users of my methods. They would essentially have to start again from scratch. Five years on were now in the process of updating and revising this analysis done back in 2007-2008. And this time I have available to me the SWISH package of data management tools. I can see at least three advantages of using this package over the way I did it back then. First of all, it's just easier to write the code. That means it's easier for me to debug. I can spend less time and effort making sure that it's correct and I can have more confidence that it's correct. Secondly, it means that I can modify the analyses. I can try out different ways of doing the analyses, doing the data preparation and the analyses. And this means that we can be more sure that we're using optimal, efficient methods for the work. And finally, and I think most importantly, it means that the methods can be communicated more easily to others. It's intrinsically better documented and it means that we can communicate the methods we used so that others can use them and perhaps improve upon them. There are five main benefits to using the SWISH system to access and merge datasets. First of all, it improves data access by providing a drag and drop interface rather than relying on complex programming. You can create executable workflows that integrate both documentation and execution. Third, workflows are easy to share for collaborative research and teaching and training. The workflows are also easy to modify and update with new data or to change the spatial scale. Lastly, SWISH workflows allow you to incorporate existing procedures and analyses and use your preferred statistical packages. The SWISH system is free open source software. You can find and download the SWISH tools and tutorials from our GitHub product site. And there's some more information about our project at our blog.