Confessions of a Data Scientist
Posted on
Data scientists are the magicians of the data world. There are a lot of skills that are required, and each data scientist will have their own speciality. For example; data visualisations, developing packages and thinking of alternative methods to get tasks completed. However, there are a handful of things which all data scientists will agree on. Below are a few examples which I experienced during my MSc at Lancaster.
Machine Learning is not the answer to everything
Unfortunately, as exciting as machine learning and artificial intelligence sounds it will not solve all of your problems. Sometimes it just overcomplicates the problem and then you have to start solving the problem again.
Get ready to see a lot of red error lines in your code
Programming codes are beautiful. They are elegant and a sophisticated way of writing mathematical algorithms. But for data scientists to get the masterpiece, which is their beautiful code, you will get a lot of errors and problems which can be as simple as a misplaced comma or as silly as the wrong depth of indentation in Python. Whatever is causing the error can take hours to figure out, but most of the time take a second to solve.
Pen and paper are still your best friend
As computer savvy as data scientists are, pen and paper are still their best friends. I am not sure if your brain can’t read the computers mind as well as it can read what is written in front of them, but when an issue arises in your code and you are getting frustrated with your program I highly recommend writing down what you want your program to do and the specific algorithms you want to do. I can almost guarantee that this will solve around 50% of all your problems.
Old fashioned graphs are the best
With more ways of representing your data becoming increasingly easy to access, they are not necessarily the answer to showing your results. Sometimes, as data scientists, we get overly excited about new graphical methods or animated graphs or just side-tracked by a graph we have never come across before, but don’t fall into the trap of using this. These graphical methods are great for the data that they have been used on, but they rarely work on the data that you have been provided.
Get to know your data
If you don’t know your data, you won’t be able to analyse it and applying new innovative techniques to your data will prove futile. Get to know your data, use summary statistics (mean, median, range, etc), draw histograms and line graphs to know what your data is showing. Everything is easier afterwards!
Related Blogs
Disclaimer
The opinions expressed by our bloggers and those providing comments are personal, and may not necessarily reflect the opinions of Lancaster University. Responsibility for the accuracy of any of the information contained within blog posts belongs to the blogger.
Back to blog listing