I had a lot of fun at the OCRUG Hackathon this weekend. Saturday night we took a break from our team presentation building to get to know each other a little while doing what we love – working with data. We were randomly assigned to teams for each of four questions. Using the tools of our choice, each team was given five minutes to make use of this penguins dataset to answer the following questions.
In having fun with the new technologies that were introduced earlier in the day, I thought it would be exciting to find out just how fantastic the Qlik Sense ‘ask a question’ insights dashboard app provides. My goal is to get the answer to these questions before my teammates using this tool.
Below are some photos of the Qlik Sense results when I simply paste in these questions and the tool that the team decided to use to deliver the answer.
Q1: Which island(s) have at least 2 penguin species?
The answer for this question was found fastest and easiest using Microsoft Excel. It was verified by a colleague using R while I was trying to figure out how to rephrase this question.
Q2: Which species has the shortest mean bill length and what is its bill length value?
Remove any missing values from the calculation.
How do I make sure that missing values were removed form the calculation? What’s the backend look like? Did I ~ oh ~
My colleagues answered this question using Excel and validated their results using R before I could feel confident that the null values were removed.
They’re null, though – I think that it can be assumed. *Head Scratch*
Q3: Create a scatter plot of bill length (x-axis) and bill depth (y-axis) colored by species.
This was handled using ggplot on R before I could begin to interpret these insights.
On a side note, Qlik Sense does do KMeans clustering, so that’s cool.
Q4: Create a scatter plot of body mass vs. flipper length for Gentoo penguins and add a linear trend line.
Put the value of the trend line slope in the title of the plot.
My colleague created the plot in R and we ran out of time before we could figure out how to put the trend line slope in the title of the plot.
Conclusion
The OCRUG hackathon hoot!
Here I shared the results that I got when directly querying the Qlik Sense application with these specific questions that were asked of the penguins dataset at the hackathon.
Qlick Sense generated some great, quick, insights; but when it comes to answering these specific types of questions – it takes a bit of work. The answers can, no doubt, be acheived; but, the questions do have to be phrased in a manner that speaks to the application.
Leave a Reply