When I first joined the federal government, the structure of the organization was deeply confusing to me. Too many acronyms, so many lines of reporting, the sheer size of the place. It's a difficult place to learn.
So I built a tool.
Today I can share it with all of you:
I wanted an easy visual resource to show me where any team is and who they report to: Give me a sense of it all. That data came from GEDS: the Government of Canada's employee phone book.
But uncompressed it's over 800MB of JSON and it's meant to describe each employee, not the structure of the organization.
So I wrote a little script to invert it and use each employee's reporting line to reverse-engineer an org chart for the whole of government. That starts to look like this:
This exercise reduces the 808MB JSON file down to a much more reasonable 8MB, or just a few hundred KB per department. WIN! Now on to visualization. I used the @MiroHQ API to build out the org chart of any department anyone wants to see.
This was fun math to get the layout right, I'm not gonna lie. But once we're in @MiroHQ we can start to zoom in on teams, link teams together and even search for teams within a broad structure.
And suddenly the entire organization of any government department is demystified, and collaborative work becomes easily visualized! FUN! The front-end lets you pick any department, and one of three layout choices (horizontal, vertical left-aligned and vertical right-aligned).
Because this is all #OpenData the code behind it is here, and that page includes instructions for how to install and use the tool: github.com/wolfewylie/gov…
And, because it's the federal government, the tool is entirely English/Francais bilingual.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
I’ve been working with the @cbcmarketplace team for almost two years building this story, building the dataset and the structures, helping the team prove how widespread these problems are. We’ve published a few stories, but this is a big one. (THREAD): cbc.ca/news/marketpla…
@cbcmarketplace The raw reports are all public information. Nothing is preventing anyone from looking them up. But it’s very difficult to look at the system as a whole. The reports live here: publicreporting.ltchomes.net/en-ca/Search_S…
@cbcmarketplace Each report is a PDF. Most of them are text PDFs, a few of them are image PDFs. They look something like this.
After the death of George Floyd, and the massive international movement to defund and demilitarize police, we got to work trying to see the Canadian context. Today, that project goes live.
There is no central database of police violence or killings in Canada. In 2018 we set out to build our own, and we launched Deadly Force, a database of everyone killed by police since Jan. 1, 2000.
555 people have been killed by police in Canada since Jan. 1, 2000. We’ve told the story before of how Black and Indigenous people are over-represented in that data. That hasn’t changed in this update.