diff options
| author | James Smith <js5@sanger.ac.uk> | 2021-04-27 12:20:26 +0100 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2021-04-27 12:20:26 +0100 |
| commit | cd96736b352012e6fc50dadc0e7759709fa31f1d (patch) | |
| tree | 1d582ed548632a68bca0c306287afbff83f9e9f9 /challenge-110 | |
| parent | 41b7248730a5f8618a9bd87edd26aecb18d548de (diff) | |
| download | perlweeklychallenge-club-cd96736b352012e6fc50dadc0e7759709fa31f1d.tar.gz perlweeklychallenge-club-cd96736b352012e6fc50dadc0e7759709fa31f1d.tar.bz2 perlweeklychallenge-club-cd96736b352012e6fc50dadc0e7759709fa31f1d.zip | |
Update README.md
Diffstat (limited to 'challenge-110')
| -rw-r--r-- | challenge-110/james-smith/README.md | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/challenge-110/james-smith/README.md b/challenge-110/james-smith/README.md index 014a7cd5d7..8e08f1b553 100644 --- a/challenge-110/james-smith/README.md +++ b/challenge-110/james-smith/README.md @@ -229,10 +229,13 @@ sub transpose_seek { * We then use the regex trick to get the first column of the data. * Memory usage: + * This script does not load the file all in one go - so really needs a lot less memory (vs more disc accesses). It is linear in the number of lines, e.g. for the 1000 line file we load in roughly 1Mb of data at a time, and the memory usage is roughly 1.3Mb. * Note this is `O(n)` as well as if the rows get longer then the number of bytes used does not increase. + * Having played a bit - the sweet spot of `$BYTES` lies somewhere between 1K and 2K. Smaller makes the + regex in the split more efficient, larger reduces the file IO overhead. ### Some information about speed/memory etc... |