Solve problems. Contrary to what people think, most geometry problems don't require much reference code and if you understand what you can extract from dot/cross product you're one observation away from solving most reasonable geometry problems in competitive programming.

At the very least, do understand the intuition behind "sweep line algorithms" and "radial sweep".

One thing to keep in mind is that most problems don't require any floating point calculation, so try to avoid bashing problems with copied formulas that you don't understand as much as possible, as you might be able to reorganize things in order to precision to not be an issue.

The most important geometry topics for ICPC are:

• 5D incremental convex hull in $$$O(n\log n)$$$ + 5D onion decomposition in $$$O(n\log^2 n)$$$

• 3D Voronoi diagram in expected $$$O(n\log n)$$$

• 4D online half-hyperspace intersection in $$$O(\log n)$$$ per query

• Fast arbitrary precision floating point numbers for super precise calculations

• Bigint as a consequence of the previous point

• FFT, Newton's method, Burnikel-Ziegler-Verfahren algorithm as a consequence of the previous point

• KD tree, Vantage point tree, R-tree for getting AC with $$$O(n^2)$$$ solutions

• HTML, CSS, JavaScript. Without these you won't be able to write visual debugging tool for all that mess

• 200+ WPM typing speed, otherwise there is no chance you will write even the first point

Have i forgot something?

use vjudge workbook : https://vjudge.net/workbook

SecondThread can suggest some topics and practice methodology for geometry problems.

Read part 1 and part 2 of Handbook of geometry for competitive programmers, it's really useful.