tldr we need to survey leetcode math and dp questions.
optional (other ticket is fine): survey the "not on algomonster graph" problem types to see if we coincidentally solve for them, or if we can get to 80%+ with only a few extra problems (less than 5).
optional: compare this list to neetcode core skills, blind 75, and neetcode 150. do any of them get over 80% or is it just a matter of "hardening" on the 67-75 range?
important schelling points for coverage are:
50%
55%
67%
75%
80%
but the algo monster pie graph itself doesn't get to 80%, so we can write that off. still, we can target "80% of the 80%" so, 64%, and indeed we can beat that so target schelling points are 67%+ (standard) and 75%+ (advanced)
we don't directly cover math, so we're 5% less than what is stated. we can review some math problems in case these are considered trivial like max sum or something and in my head they are closer to general programming than math. but we don't do any fancy math like the knight chess board math, so we will want to
tree algos + caching are a significant subcategory of DP*, so you will get some of this even though we don't formally target it.
*I think DFS + Memoization is a kind of top-down DP
So, "tree algos with caching" is a subcategory of DP, but do we know anything about how large this subcategory is?
Maybe all of them, as subproblem connections can always be viewed as graph edges?
In an interview context do we have a sense for whether this is a majority or minority of DP questions?
TODO: brute force analyze top 100 leetcode DP and determine how many are tree algo + caching
we can also make a contingent statement like "if you view subproblem relations as graph edges, then..."
https://www.ladderly.io/blog/2023-11-25-leetcode-kata#calculating-estimated-coverage
tldr we need to survey leetcode math and dp questions. optional (other ticket is fine): survey the "not on algomonster graph" problem types to see if we coincidentally solve for them, or if we can get to 80%+ with only a few extra problems (less than 5). optional: compare this list to neetcode core skills, blind 75, and neetcode 150. do any of them get over 80% or is it just a matter of "hardening" on the 67-75 range?
important schelling points for coverage are:
but the algo monster pie graph itself doesn't get to 80%, so we can write that off. still, we can target "80% of the 80%" so, 64%, and indeed we can beat that so target schelling points are 67%+ (standard) and 75%+ (advanced)
*I think DFS + Memoization is a kind of top-down DP So, "tree algos with caching" is a subcategory of DP, but do we know anything about how large this subcategory is? Maybe all of them, as subproblem connections can always be viewed as graph edges? In an interview context do we have a sense for whether this is a majority or minority of DP questions? TODO: brute force analyze top 100 leetcode DP and determine how many are tree algo + caching we can also make a contingent statement like "if you view subproblem relations as graph edges, then..."