Top 4 Foundational Skills for Data Scientists

The study of statistics is a thousand years old. Artificial intelligence and data mining have been around for decades. The terms 'data science' and 'data scientist' are very young being coined in the early 2000s. Today, actionable data has become a critical decision-making tool for businesses of all sizes. The job market reflects data's increasing importance. LinkedIn has over 7,000 data scientist job postings. Data Scientist has the top spot in Glassdoor's 50 Best Jobs in America report.

Job postings list many skills such as SAS, R, statistics, supervised learning and many others. However, few applicants will possess all the listed skills. This is due to the relatively young age of data science as a discipline of study and its still evolving role. There are four things that most data scientists should possess. They need to be proficient in the programming language Python, knowledgeable about data frameworks, well versed in Structured Query Language (SQL) and be a good communicator.

A data scientist has many sources of information. There's internal data like company sales and customer statistics. External data can come from social media, government data, consultants or vendors. Rarely is any of this data in a format immediately useful to a data scientist. The task to clean and transform the data falls on the data scientist. Enter Python.

Python has built-in data structures and dynamic semantics making it an ideal tool for data scientists. One of Python's code libraries is called the Pandas library. The library has pre-built, ready to use structures that make quick work of data analysis and transformation. It is the third most desirable and useful skill for a data scientist.

Data scientists work with very large amounts of data. Storing, manipulating and analyzing that data is done on a hardware platform designed for exceeding big data requirements. These platforms are composed of multiple servers sharing the processing load. While the scientist is not responsible for maintaining the hardware, they must know the software framework that controls data storage, configuration and access.

Hadoop clusters are one example of a distributed data environment. Hadoop is a collection of open source applications that make big data analysis manageable. The framework runs within a cluster of servers. Because Hadoop is open source with cost-effective scalability, it has found uses in businesses of all sizes and industries. For working scientists, managing a Hadoop setup is the second most important skill. Whether it's a Hadoop or Spark framework, a data scientist must be a proficient data manager.

SQL is the language of relational databases. So widespread is its use that there are multiple variants of SQL depending on the database computing environment. Enterprise databases like Oracle or Microsoft SQL server have proprietary implementations of SQL. Software vendors have incorporated SQL into their products. For example, Apache's HIVE is a data warehouse application that facilitates dataset access and management using SQL. SQL is ubiquitous. It should have a permanent place in any data scientist's skill set.

Google's chief economist Dr. Hal R. Varian stated in a 2009 interview, "The ability to take data—to be able to understand it, to process it, to extract value from it, to visualize it, to communicate it—that’s going to be a hugely important skill in the next decades." Communication is even more critical today. Business intelligence tools are improving leading to an increase in the quantity of insights and information generated. Insights must be communicated if they're to be actionable. Numbers by themselves are only numbers. The scientist must provide commentary and context. They must clearly explain and articulate the significance of their findings.

Learning Python, SQL, data frameworks and becoming a better communicator are desirable skills for those entering or transitioning to data science. It's inevitable that as data science evolves, a data scientist's skillset must also expand. A solid foundation can ensure future success.