A Week's Reflection at the Data School: Looking back at my application

by Anne Porcia Affi-Asamani

As I approach the end of my first week of training at the Data School, I couldn’t help but compare the foundational data concepts we've covered so far to my previous experiences using data, especially in my application process. I’ve found that several things that I encountered and attempted while creating my viz had simple solutions; and that even with one week of training, making the vizualisations I did for the application could have taken half the time if I knew then what I knew now (and if I had access Tableau Prep).

Here are two examples of not only techniques I already used that I now know the names of but also techniques that could have simplified my process:

  1. Joins/Unions

Joins and Unions involve combining more than one data set either through joining more than one table together (Joins) or by a stacking one data set on top of another (Unions). To join two tables they must have one data field in common while unions must have all of their data fields in common.

In the Viz that I created for my final interview I needed to analyse data from the London Fire Brigade. I joined census data to the LFB data to supplement my analysis.

  1. Splitting Columns

Splitting columns involves separating columns in a data set into two separate data fields showing a difference in data types, measures or categories.

In the LFB data, there were values in the data set labeled 'False alarm - good intent' and 'False alarm - Malicious' which described whether or not a specific incident was a false alarm and the nature of that false alarm. I could have created a column for false alarms and split the column to do a further analysis into the kind of incidents that occur in each borough I included. (This would have also involve pivoting which is another technique I could have used.)

Anne Porcia Affi-Asamani