Python

Python is the default language of data analysis and machine learning, and a reasonable first language for someone moving into engineering from another field, because it reads closer to English than the alternatives and gets to a working result quickly.

What the work consists of

In data roles: loading messy data, reshaping it, and producing something that answers a question — overwhelmingly with pandas rather than with the machine learning libraries that get the attention. In engineering roles: the same work any language does, inside somebody else’s codebase.

What employers screen for

For data work, whether you can explain what a transformation did to the data rather than only that it ran without error. A notebook that produces a number nobody can trace is worth less than a short script whose steps are legible.

• pandas at working depth: merges, groupby, and what happens to your row count at each step
• Writing a script someone else can run next month without you present
• Virtual environments and dependencies, which is where self-taught Python most often shows its gaps

The honest limit

Python is rarely the thing that gets a junior analyst hired in this market. SQL and a spreadsheet will get you into most entry-level analyst roles, and they are what the interview tests. Python becomes valuable at the point where the analysis you want to run genuinely outgrows a spreadsheet — a real point, but one that arrives later than course marketing implies. Learn it after SQL, not instead of it.