For day two of dashboard week we've been asked to create a visualisation in Power BI using data that we web scrape from basketball-reference.com. We've also been asked to make sure that we include at least 20 URLs within our web scrape and can provide any analysis we like.
I really wanted to make my analysis more focused today and challenge myself to tackle a large dataset. I, therefore, chose to create a dashboard looking at the active coaches in the NBA over their entire careers. This required me to input around 90 tables from 30 URLs. Although web scraping from a single URL in Power BI is simple when you start inputting more URLs the software becomes really difficult to use. Each time you want to run a query to do a simple thing like change a field type each URL appeared to also be queried. Ultimately I managed to get all the tables I needed and cleaned them however, I was left with very little time to create my dashboard.
I had made a detailed sketch of the dashboard wanted to create whilst waiting for Power BI to query however when I started building I encountered several problems. I planned to create a Sankey chart and knew this was possible in Power BI but unfortunately, you can only include a maximum of 8 categories within the chart. I, therefore, replanned my dashboard and simplified it to make sure I had something that was finished. In the end, I was fairly pleased with my dashboard however there were some simple mistakes that I would know how to fix in Tableau. I really struggled with editing filter actions and definitely need to practice working with Power BI some more.