Opinion
Data Science started to become a thing in my life in 2019. Since then, I have been learning data science passionately and curiously, discovering something new almost every day.
Data has always been the center of data science, which is the normal to expect. Data science is all about extracting a piece of information, insights, or value from the data.
The potential of creating business value with data has increased noticeably, which has driven enormous demand for data, tools, and skills. As a result of all these transformations, the number of tools skyrocketed.
At the beginning of my data science journey, my understanding of being a data scientist is a person who is able to understand what data tells us. Data scientists go beyond the superficial and extract the sort of hidden pieces of information.
As of now, I feel like being a data scientist is more associated with how good you are at using tools. Don’t get me wrong. Your ability to read, understand, and make sense of data are still of crucial importance. However, your skills at certain tools have come into prominence.
My understanding of data science is becoming more like how good you are at using the tools in the data science ecosystem. I’m not sure how to feel about this.
We cannot ignore the tools, obviously. It has always been imperative to use software tools and packages to clean, process, and analyze data because we are likely to deal with large amounts of it. However, the quantity and complexity of the tools have been increasing. As a result, data scientists spend a great deal of time learning to use such tools.
What is wrong with this?
First of all, there is nothing wrong with using more efficient and faster tools. Time and computation power are significant resources so if there is a tool that has the potential to save us some time or computation, it is not wise to ignore it.
However, this should not cause you to ignore what is more important: Data.
Having a good grasp of the data at hand can do much more than any other tool. For instance, there are several hyperparameter tuning tools, which can help you find the optimum hyperparameter values and possibly achieve an improvement on the model.
On the other hand, deriving a new feature by mastering the structure and properties of the data and getting some business context included, you might well achieve a significant improvement compared to what hyperparameter tuning offers.
Do I have to learn all the tools?
Absolutely not! However, data science is not a solo-practitioner kind of job. You work as a part of a team so the tools used by the team are what you need to learn. Considering the high turnover rate in the data science ecosystem, you are likely to change companies frequently, which might mean new tools to learn.
Things might be slightly different for large corporations
If you work as a data scientist at a large corporate company, your tasks might be limited to a narrower focus. In such cases, you won’t have to deal with or learn so many tools because you do focus on a particular task.
I think this is only valid for a limited number of companies. A substantial amount of data scientist jobs require you to participate in multiple steps of the data science or Machine Learning workflow.
What I am trying to explain comes down to being a data scientist vs being a data science tool expert. You can’t ignore either one of them completely. However, I strongly recommend not losing focus and not forgetting about what data science is really about.
If you become a tool expert, you can perform the tasks you are told to do. This is definitely a valuable skill. If, on top of this, you help make decisions with data, your chance of becoming an outstanding data scientist tremendously increases.
Know your data well!
You can become a Medium member to unlock full access to my writing, plus the rest of Medium. If you already are, don’t forget to subscribe if you’d like to get an email whenever I publish a new article.
Thank you for reading. Please let me know if you have any feedback.
