Core Python for data work

Before any library, plain Python needs to feel automatic. That means variables, lists, dictionaries, tuples, loops, functions, comprehensions (one-line loops that build a list), f-strings, imports and reading a stack trace, the error report Python prints when something breaks.

Every NumPy and pandas object is built on these ideas, and their error messages assume you know them. Write small functions that take data in and hand data back, rather than long scripts that change global variables. That habit is what lets you rerun your cleaning code later.

Also learn virtual environments (venv or uv) and pip, so each project keeps its own library versions.