Data science is a complex field, requiring expertise in several subjects including statistics, mathematics, machine learning algorithms, data visualization, and various tools like Hadoop. It is easily the most sophisticated field, requiring advanced skills and strong keenness toward problem-solving. Among these skills, whether coding is important is as hotly debated as any other pressing global concerns.
Unfortunately, there’s no direct answer to the question. The industry, experts, and individuals are divided on the answer. While a group of industry experts strongly believes that coding is a key skill to master data science, the industry differs in its opinion. What creates more confusion—the data science certification course and program consider proficiency in R, Python, Java, and SQL as an important parameter to gauge the competency of data scientists.
Let’s understand the views of the experts.
The No Squad
This industry squad believes that coding is a critical skill that decides whether a candidate is a good data scientist. Eric Hulbert, popular known as the ‘analytics dude’, answers the question with a single ‘nope’. He explains the nuances of coding in a data scientist’s work and corroborates his point with personal experiences. Similarly, Racheal Tatman, who writes for Free Code Camp, explains that to emerge successfully in data science, you should be able to write code for statistical computing and machine learning.
Ronald Van Loon shares data scientist should possess knowledge of programming languages like Python, Perl, C/C++, SQL, and Java. He adds a data scientist should have expertise in tools like SAS, Hadoop, Spark, Hive, and Pig.
Burtch Works, a leading executive recruitment company, shares ‘Python (along with Java, Perl, C/C++) and machine learning’ are must-have skills that employers are looking for. Experience in Hadoop, Hive, or Pig is a strong selling point, the company adds.
Several universities also belong to the no squad. A few universities including Columbia University even consider that programming is an essential part, while statistics are not so much. This is way worse. With today’s ever-changing DS realm, one could argue that statistics is one of the most important parts of data science. In 2021, it’s certainly not the least important.
The Yes Squad
A fraction of the industry believes that one can become a data scientist without knowledge of programming. There are valid reasons as to why the experts believe this. The following are a few of them.
• Common machine learning algorithms are known and optimized.
• Trifecta and Tableau are drag and drop interfaces. Similarly, other applications allow data decisions based on items you can use.
• Automation is slowly taking over data science. Google’s Cloud AutoML and DataRobot allow data science professionals to find the right algorithms. You can train high-quality custom machine learning models with minimum effort and machine learning expertise.
• Google Duplex demo further automates decision making. It will allow data scientists to have a conversation with a machine to make decisions rather than coding.
Increase credibility
In addition to coding, a data scientist requires a gamut of skills. As data science is new in the industry, a globally-recognized data science Python certification course is a way to acquire skills and proof to demonstrate skills. DASCA’s ABDATM (Associate Big Data Analyst), SBDATM (Senior Big Data Scientist), and SDSTM (Senior Data Scientist) are prominent certifications that prove the holder’s overall competence as a data science professional including proficiency in R and Python.
Similarly, IBM’s Data Science Professional Certificate and Dell EMC Proven Data Science Professional prove candidate’s competence as a data science professional including coding, knowledge of machine learning algorithms, statistics, and other requisite skills. The above-mentioned certifications are also the industry’s best certification for data science professionals, perfect for aspirants looking to acquire skills.
