New day, new tool ! Today, Andy has challenged us to use Power BI to produce a dashboard on data from New Orleans.
So I started by getting the data from the NOLA 311 API, cleaned it on Alteryx, changed the format for a more appropriate one – or so I figured – to be able to work on those data later, I then watched a few short tutorials about Power BI ( some are available here and here) and OH MY GOD IS IT 12PM ALREADY??? How come the time flies by when you would need it to get slower…
Anyway, my point of view about Power BI: it is interesting when you want to throw quickly some data and make a quick chart, along a meeting for example. With a few clicks only, you can generate a graph. But as soon as you want to get a bit more granular and combine different measures.
But I must say I felt a bit frustrated. It is like having to learn again how to walk when you started running for the past 3 months…!
I have tried several things today, my objective was to test the features of the tool and see what I could build. It was quick indeed, but as soon as I wanted to create something a bit more customized, it was a pain. Power BI is a free tool that thus has some serious limitations. The automatic filtering feature soon becomes irritating and you really want to compartmentalize your pages and narrow the topic.
The simplest thing becomes quite complex, like creating new measure. For example, I tried to type the equivalent of a calculated field, a percent of total from a count distinct. This is what I have found on the community forum:
So what does my dashboard look like ? Well, it won’t be my masterpiece, but I am happy to share what I have learnt today!
My daily self assessment:
- What to keep doing for the next time: Asking for help before time flies by. I got lost in the APIs documentation and saw every other DSers starting to prep the data while I was still trying to figure out where to fnd the name of the key for the token. Not to even mention the necessity to address the limit in downloading the data.
- What to improve: The use of Pomodoro. I mentioned this one in the daily post, and then, I did it again: rushed into the data set without checking the time….
- What to stop doing Well, I didn’t do any particular thing that I fundamentally judge bad, except for that delay in publishing my article (Blame Power BI online platform for that, I spent ages finding my work online !)
Now, tomorrow is a new day, let’s see what Andy will have prepared for us !