Inadvertently saw the difference constraint from the introduction to the algorithm and spent an hour simply learning
Reference blog: Blog 1
It probably means that there are generally two models for this type of problem:
1. Give you n intervals to indicate the maximum number of intervals, then find the total maximum?
2. Give you n intervals to indicate the minimum number of each interval, and then find the total minimum
Example: P1250 tree planting
Topic:
Give you m intervals bet means that at least t trees are planted from b to e, and then find the minimum total trees.
Analysis formula:
Because at least t trees are planted, it must be written in a form greater than or equal to
s [e] -s [b-1]> = t // Only lower bound, no upper bound
Each position i can be planted or not: 0 <= s [i] -s [i-1] <= 1
s[i]-s[i-1]>=0
s[i-1]-s[i]>=1
Building plans:
b-1 to e Directed edges with edge weight t
i-1 to i Directed edges with edge weight 0
As for why i to i-1 build a directed edge of -1
Personal understanding is that the above two sides are established by the minimum value of this side, then i to i-1 should also be the minimum weight (the value is 0,1)
Another question: P3084 Spotted Cow
Topic:
Give you m intervals (l, r) means there is only one cow in interval l, r. How many cows are there in total?
Analysis formula:
s[r]-s[l-1]=1
Each position i has a cow and no cow: 0 <= s [i] -s [i-1] <= 1
Written in the form of the last edition only
s[i]-s[i-1]<=1
s[i-1]-s[i]<=0
Building plans:
Directed graph with weights of 1 from 1 to r
r directed graph with l-1 construction weight of 1
Directed graph with i-1 to 1 building weight value of 1
Directed graph with i to i-1 building weight value of 0
Here is the same, each side is to save the maximum value, and then run the shortest path on the graph.
Because of the negative weights in the graph, Bellman-Ford has only learned SPFA solutions.