“The Adventure of Pinocchio” is the most translated book (non-religious). It was written in 1883 by the italian Carlo Collodi (sorry, it was not created by Disney) and it has been translated to more than 250 languages. This means that the story about Pinocchio is more meaningful than the actual language it was written on. Even the most popular version is not in his original language.
Fast-forward to a couple of days ago and while trying to fit a perceptron using Python, I realize that I am investing more time and effort trying to get the machine to work than analyzing the actual problem. I’m focusing more in the language than on my Pinocchio’s story. This is particularly bad because my Pinocchio story is also going to be translated into several languages (finance, marketing, life) and the readers are not going to care about the original language but only about how the story is going to affect them.
Don’t get me wrong, knowing how to use different tools is very important but understanding the analysis is paramount. During my undergraduate degree I use Stata to learn econometrics. That was the first and last time that I used Stata. On the other hand, I use Econometrics all the time throughout all sort of software and programming languages; but the point of analysis is always the same.
More recently I have been fitting Machine Learning and Neural Network algorithms using both R and Python. Even though I prefer R (calm down serpentine friends), I must say that in a couple of years, it might be irrelevant as more and more machine learning algorithms are going to be fully integrated on software or APIs, rendering the application of them as trivial as adding two cells in Excel. In fact, Excel has an adding to connect to all the Machine Learning capabilities of Azure.
Everyday we are bombarded by a new app, software, programming language that it’s going to make our lives easier. Alteryx, Knime, RapidMiner are the hot words now and they are fantastic: Alteryx streamline your data cleaning and manipulation process; Knime is the new open source platform for data mining and machine learning and RapidMiner automates the machine learning pipeline. If you think they would increase your productivity, by all means, learn about these programs. But never forget that these are only tools and not the actual story. Is better for you to focus on the story; your Pinocchio is going to be translated to other languages anyway.