search for duplicates

RSLogix 5000

Need to search a 600 point string array for matching numbers, eg: SPRD(0) to SPRD(599)

If I have 2 part numbers the same then set a bit high.

Do a compare and index both inputs. Index thefirst from 0 to 1,2,3 etc. When you reach 600 infrx the second from 0to 1 and rerun the 1st instance. Once complete index from 1to 2

david.hinch said:

RSLogix 5000

Need to search a 600 point string array for matching numbers, eg: SPRD(0) to SPRD(599)

If I have 2 part numbers the same then set a bit high.

Click to expand...

Could these part numbers be stored as DINT values, or too many digits and/or embedded non-digits?

If you can put them in a DINT array, it would be more efficient to sort and then scan the array once; not to mention a DINT compare vs. string compare.

If not, then carefully consider whether the nested looping should be done all in one scan, or spread over multiples.

Fsc

You can use FSC to compare a array against itself with a bit of work. I had to make sure the .POS and the Index tag weren't the same else you would find a match since you would be comparing itself against itself.

If you increase the .POS by 1 past the index you don't get a match and it would speed it up so your not re comparing data for a second time. [3 vs 5] then [5 vs 3] later on.

This way the first compare is [0 vs 1-19] and then the next compare is [1 vs 2-19] etc. till [17 vs 18-19] etc. I did a 20 array so I wouldn't have to hand enter data into 600 array.

Sort a copy of the array an then look for a code repeating.

Mere said:

You can use FSC to compare a array against itself with a bit of work. I had to make sure the .POS and the Index tag weren't the same else you would find a match since you would be comparing itself against itself.

Click to expand...

This clever approach has the feature of working over multiple scans. Therefore it might be advisable to copy the string array into a working copy for the FSC before starting, in case the base array could change while duplicate-checking. Otherwise prevent any changes while this logic is running.

One thing I did not see, and it could be below the rungs in the screen shot, is handling the situation where there are no duplicates. It appears that the processor would eventually fault on an out-of-bounds index in this case.

The original code had a move zero to index when it reached the size of the array. That had gotten dropped out when I built this. Also how I had this set up it stops once it finds its first match. You can have it pause and move the .pos and index to a array to create a list of matches. Then unlatch the required tag and have it start back up.