I am trying to learn 1 look PLL, but I don't know where to start memorizing. Should I start with number 1 and move on, or should I learn certain algorithms first
I learned the edge cycles, one that swaps 2 adjascent corners (J), and one that swaps 2 opposite corners (Y). Later on I learned the remaining easy ones (A, T, H, Z, R). It worked out pretty well.
I suggest you save the hardest ones for last (N, G, E etc.).
I learnt the 2-look PLLs first, then the rest in roughly the order they are on badmephisto's PLL site.
I left the hardest ones until last, Ns and Gs. N's were not as hard as I thought. Gs took ages to fully recognise and execute right each time. I don't like the E perm either.
Some PLL cases come up more often, so you could learn them first. The chance of getting one of the 4 G-perms are the most likely by numbers, and it's not that hard to learn. I learned them in "pairs" where i practised by doing a G-perm and it's inverse and then moved on to practice the mirrored.
Also when you learn the V-perm you also almost instantly learn the F and you when you learn the Y-perm you also learn 2 OLL's.
Badmephisto has a very nice guide with tips for memorisation and recognition, and both Yu Nakajima and Erik Akkersdijk has some very nice videos of their algorithms and how they perform them.
I'm working on a PLL Guide for you. It's going to look similar to my OLL Guide. You should learn them in the order listed. I highlighted the cases you need for 2 look.