We recently partnered with the team at the South Pacific Regional Fisheries Management Organisation (SPRFMO) to help streamline their reporting.
SPRFMO reporting
SPRFMO approached Dragonfly Data Science to help rebuild its database of fisheries data in the South Pacific Ocean, and produce its annual summary reports for the scientific committee meetings.
Executive Secretary at SPRFMO, Craig Loveridge, says Dragonfly opened their eyes to automating a lot of their reporting features.
“It really did make our work more efficient and save us time. That was a big deal.”
“We collect a lot of information, but sometimes getting that information out in a usable and readable format can be difficult. The reports that Dragonfly helped develop meant we could display and share that information widely, making it more valuable to users.”
He says most importantly, Dragonfly respects data confidentiality and sensitivity which is integral for SPRFMO and their partners.
Writing code to produce reports
Dragonfly Data Scientist Hayley Wikeepa says producing a data intensive report doesn’t need to be a painstaking manual process. By writing code to carry out the analysis, the process can be made efficient.
The three key components of building a good report are accurate data, thoughtful analysis and discussion, and presenting it all in an accessible format.
“SPRFMO’s reports includes things like what fish were caught, where they were caught, and who was involved. Writing code that consistently analyses this information allows us to help automate the report creation.”
“It takes a bit of thinking and work up-front to collate datasets, get the formatting right, and ensure the first report is consistent and secure, but once you’ve got one report done, it’s much easier to reproduce year on year,” she says.
Dragonfly scientist, Katrin Berkenbusch, says many organisations will write reports by manually editing excel spreadsheets, generating graphs, and transferring data into a presentation format like a Word or Google document.
“But, doing things manually like this means it’s easy for errors to sneak in because you’re having to update data in multiple places manually, particularly when dealing with large, complex datasets. Reproducible data methods significantly reduce that risk of error.”
“With repeatable methods, the final report is built from the underlying data, so any updates to the data flow through to the report,” she says.
Talking the same language
The SPRFMO team enjoyed working with Dragonfly.
“It was fun, you know, they were excited about the information, excited about the work and it’s nice to be able to talk to people in the same language,” says Loveridge.
“To be able to partner with somebody who has similar goals as you have makes such a difference.”