Talk:Nonogram solver: Difference between revisions
Content added Content deleted
| Line 3: | Line 3: | ||
: It's a wiki, posted code is not my private property. Do whatever you think is right. --[[User:Ledrug|Ledrug]] ([[User talk:Ledrug|talk]]) 20:22, 31 March 2014 (UTC) |
: It's a wiki, posted code is not my private property. Do whatever you think is right. --[[User:Ledrug|Ledrug]] ([[User talk:Ledrug|talk]]) 20:22, 31 March 2014 (UTC) |
||
:: OK, I have updated the Python entry. But we are often collaborating on RosettaCode, so it's very good to be civil and friendly :-) I don't want to step on your toes. -[[User:Bearophile|bearophile]] ([[User talk:Bearophile|talk]]) |
:: OK, I have updated the Python entry. But we are often collaborating on RosettaCode, so it's very good to be civil and friendly :-) I don't want to step on your toes. -[[User:Bearophile|bearophile]] ([[User talk:Bearophile|talk]]) |
||
:: Most of the run time of the D entry is used by genRow function. (And generating unsigned bytes instead of ints reduces the memory a lot, but leaves the run-time unchanged). So is is possible to use a 32/64 unsigned integer to represent the inner arrays (the possible configurations for each row), something like [2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2] => 81fff? And if a more compact representation is possible, is it leading to a faster program? (if you use 64 bits you can't have rows longer than 64) -[[User:Bearophile|bearophile]] ([[User talk:Bearophile|talk]]) |
|||
Revision as of 20:33, 6 April 2014
Python Entry
Many details in the Python entry are not pythonic (or not efficient, or both), do you accept suggestions or a link to a fixed version? -bearophile (talk)
- It's a wiki, posted code is not my private property. Do whatever you think is right. --Ledrug (talk) 20:22, 31 March 2014 (UTC)
- OK, I have updated the Python entry. But we are often collaborating on RosettaCode, so it's very good to be civil and friendly :-) I don't want to step on your toes. -bearophile (talk)
- Most of the run time of the D entry is used by genRow function. (And generating unsigned bytes instead of ints reduces the memory a lot, but leaves the run-time unchanged). So is is possible to use a 32/64 unsigned integer to represent the inner arrays (the possible configurations for each row), something like [2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2] => 81fff? And if a more compact representation is possible, is it leading to a faster program? (if you use 64 bits you can't have rows longer than 64) -bearophile (talk)