Data structures courses often front-load abstract theory before any concrete payoff. Here is a more practical ordering.
Start with arrays and hash maps
These solve the overwhelming majority of real problems you will encounter, and understanding their tradeoffs (lookup speed vs ordering) pays off immediately.
Then trees, briefly
You do not need to hand-implement a red-black tree. You do need to understand why a sorted structure lets you search in log time instead of scanning everything.
Save graphs for when you need them
Graph algorithms are powerful but genuinely rare in day-to-day work outside specific domains. Learn them when a real problem calls for it, not preemptively.