Last week DS 13 has got a break from client project, so our Friday activity consisted of a dashboard project, as it has been during the first month of Data School.

The project consisted of analysing the data that Andy downloaded from his Strava Account. Over 1400 sports activities during a 7 years time-span.

The most challenging part of this project consisted of crafting a useful dataset for the actual analysis in Tableau. The activities log were stored in .gpx files, a format not natively available as input in Alteryx. To read and export these files data in Tableau, we had to treat them as .xml files, in order to extract all the data from them. Due to this, we all ended up to create quite intricates workflows.

Since Andy requested us to use specific metrics (distance, speed, pace) I have built a workflow that would deliver all of them.

What I did not consider well enough was the timing. It took me roughly 3 hours to complete the workflow, leaving only one hour for the actual dashboard creation. On top of that, I have to admit that I’ve not fully planned my work before the start. I was so worried about extrapolating data from these files that I forgot to adapt my workflow to my final dashboard aim.

If I would have started with a clear picture of what to visualise in Tableau, I probably could have skipped several steps in Alteryx, saving a conspicuous amount of time.

So, this is the lesson learned this week. Not only the dashboard needs to be sketched, but the data prep process also needs to be planned thoroughly. Data prep, especially on large files can be really time-consuming, hence cutting down these times becomes crucial while working on a restricted time schedule.

On the right side, I’m happy that I managed to produce an effective Alteryx workflow, with minimal help. I have spent quite a lot of time lately practising Alteryx, so I’m glad to see the outcomes of this.