suzanne.soy/tock-mirror
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Corrected some of my documentation of flowAlgorithm that has repeatedly confused me ever since I wrote it

Browse Source
This commit is contained in:
Neil Brown 2008-11-20 15:39:06 +00:00
parent e117422c31
commit be6c57ce5b
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
flow/FlowAlgorithms.hs
View File

@ -26,7 +26,8 @@ import qualified Data.Set as Set
import Utils import Utils
data GraphFuncs n e result = GF { data GraphFuncs n e result = GF {
-- (Node, Edge) -> previous assumed input -> current aggregate effect -> new aggregate effect -- (Node, Edge) -> current value for said node -> current aggregate effect for
-- the current node -> new aggregate effect
-- If it is the first node to be processed for this iteration, it will be -- If it is the first node to be processed for this iteration, it will be
-- given Nothing, otherwise the result is fed back when processing the next -- given Nothing, otherwise the result is fed back when processing the next
-- node. The second parameter is from the last iteration. -- node. The second parameter is from the last iteration.
@ -63,8 +64,8 @@ data GraphFuncs n e result = GF {
-- --
-- The general idea of iterative data-flow is that all nodes start out with -- The general idea of iterative data-flow is that all nodes start out with
-- a default "guessed" value. Then each node is processed in turn by using -- a default "guessed" value. Then each node is processed in turn by using
-- the previous value (to start with, the default value), and the values of -- an accumulating value (to start with Nothing), and the values associated with
-- all nodesToProcess in the graph. This algorithm is performed repeatedly, -- each of the nodesToProcess in the graph. This algorithm is performed repeatedly,
-- processing all nodes, and if a node changes its value, re-adding all its -- processing all nodes, and if a node changes its value, re-adding all its
-- nodes in the other direction (nodesToReAdd) to the worklist to be processed again. -- nodes in the other direction (nodesToReAdd) to the worklist to be processed again.
-- --