In 1980 I made a program for passing the diagram of activities for a large 
project, computing duration and cost of the project, based on durations and 
costs of the activities. 
In PERT you have a record per node and a record per edge in the graph. I didn't 
do it that way. I still had a record per node but no record describing the 
edges. Rather the numbering of the nodes defined the edges.
This numbering is documented here. ORDINAL FRACTIONS - the algebra of data
  
|  
|   
|   
|   |    |

   |

  |
|  
|    |  
ORDINAL FRACTIONS - the algebra of data
 This paper was submitted to the 10th World Computer Congress, IFIP 1986 
conference, but rejected by the referee....  |   |

  |

  |

 
The numbering of activities is such that activities 1 and 2 are in series and 
11 and 12 are in parallel and so on. I used 11-figure ordinal fractions where 
the odd-numbered digit positions indicated series and even-numbered digit 
positions indicated parallel.
Of cause the durations of activities in series must be added together, (+/) and 
durations of activities in parallel must be maxed together (>./).
This is the smarter way to handle activity diagrams. 
Thanks.
Bo  

    Den 3:31 onsdag den 29. november 2017 skrev Daniel Lyons 
<[email protected]>:
 

 

> On Nov 28, 2017, at 7:11 PM, Henry Rich <[email protected]> wrote:
> 
> You're right about the last bit - you can't start with a transitive closure.  
> I missed that.
> 
> Suppose you turn each dependency into a J sentence, say
> 
> a =: 5
> 
> or
> 
> c =: longerof (a , b)
> 
> or
> 
> f =: a following (e , g)
> 
> or combinations thereof.  There must be no loops in the dependencies, so you 
> could do a topological sort on the tasks, to ensure that each assignment 
> refers only to names that have already defined.  Then just execute the sorted 
> script and see the results.
> 
> You would need the definitions
> 
> longerof =: >./
> following =: +
> 
> and you need a topological sort, which isn't too hard.  I know I wrote one 
> for the Advent of Code problems a couple of years ago.


That's kind of a cool idea. Is it commonly done, generating a block of code in 
a string and evaluating it?

-- 
Daniel Lyons




----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

   
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to