I love cycling, both watching and doing it myself. One of the cool things is that is very data driven sport and thanks to Strava much of this becomes available to anyone, not just the elite athletes. Of course I would love to do more and get this data into Power BI. Thanks to Strava’s API’s and the new Power BI custom data connector SDK this is now possible!
I started building this connector the day the SDK became available but unfortunately got stuck a bit and I wasn’t happy with the code I wrote, I knew this could be done easier but I am a DAX guy not an M guy at the same level as DAX :). But thanks to the new amazing examples Matt has created I was able to finish the connector!
To start, the connector I wrote is fully available on GitHub: https://github.com/Kjonge/PowerBIStravaConnector. It is mostly based on the samples provided by the data connector team so I won’t go too much into the actual building of the connector itself. I added many comments to the code and point to the samples I stole them from . But honestly it is pretty straightforward copy and paste from the samples when your service uses REST API’s with OAuth 2 and JSON payloads.
So let’s look at how to get it running yourself before we look at using it:
- First make sure you have the latest version of Power BI desktop
- Second make sure you have Visual Studio (2015 or 2017) installed so you can download and set up the Power Query SDK from here: https://marketplace.visualstudio.com/items?itemName=Dakahn.PowerQuerySDK
- Next you need to get the code from my project from GitHub, just go to https://github.com/Kjonge/PowerBIStravaConnector and click on download and open in Visual Studio:
- This now opens it in Visual Studio where you can “clone” the code to your local HD
- This opens the project where I can double click the solution
- Now this will open the solution that we need:
- As you might notice two files are missing, these are the files that are needed to authenticate with Strava using OAuth, a Client ID and Client Secret. In the current version of the Connector SDK this information is needed locally to be able to connect. There is an item on the road map (before it goes RTM) to move this into a cloud based solution so you can distribute connectors without having to add the client id and secret into the code itself. But for now during the preview we need to add them locally, so I add these files for now.
- To get the Client ID and Secret I need to create an app with Strava, so I go to https://www.strava.com/settings/api to create the app. This will allow me to interact with strava information on a users behalf. I go to my profile and “API application”
As you can see (well I deleted my Client ID and secret) but now you can get the client ID and Secret
- Also important is to set the right redirect URL, in this case OAuth.PowerBI.com, this used in the OAuth Authentication flow:
- I use that information to add two files to my solution called client_id and client_secret that contain just those 2 values that you can copy from the app you just created in the Strava site.
- Now you will notice two new icons next to the files, this means these files are excluded from github so you will never accidentally publish the client id and secret when making updates to the connector (feel free to :))
Also it is important that both these files are set to Compile so that they will be included in the build (they should be automatically but it is worth double checking)
- Now we are ready to build the connector and load it into Power BI desktop. Just Build the solution and you should get a compiled mez file in your debug folder (it doesn’t matter if you do a release or debug build)
- This file I copy to a folder that the Power BI desktop will look at “C:\Users\YOURUSERNAME\Documents\Microsoft Power BI Desktop\Custom Connectors”
- Next I have to make sure I turn on the Preview feature in Power BI desktop:
- If this is turned on I should now get the Strava connector available.
Now that we have the connector available, let’s look at using it:
- I now have a Strava connector in my Power BI desktop
- When I select it I get asked to sign in:
- Clicking sign allows me to sign in to Strava and Connect
- After connecting it shows me the Navigator where the real magic happens:
- Here you see 2 tables that are exposed, 1 with Activities and another with Athlete information, I select both. Clicking on them gives me a preview:
- After pressing Load the data gets loaded into Power BI desktop. Done!
- This now gives me a ton of data for all my tracked activities in Strava:
- That I can now use to analyze:
- And build a report on: