| --- Log opened Wed Nov 21 00:00:24 2018 |
| 00:52 | | celmin|sleep is now known as celticminstrel |
| 02:03 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has quit [Ping timeout: 121 seconds] |
| 02:05 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has joined #code |
| 02:05 | | mode/#code [+ao Reiver Reiver] by ChanServ |
| 02:08 | | Derakon is now known as Derakon[AFK] |
| 02:23 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has quit [[NS] Quit: No Ping reply in 180 seconds.] |
| 02:23 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has joined #code |
| 02:23 | | mode/#code [+ao Reiver Reiver] by ChanServ |
| 04:06 | | Derakon[AFK] is now known as Derakon |
| 04:37 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 04:37 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 04:59 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds] |
| 05:16 | | Derakon is now known as Derakon[AFK] |
| 05:16 | | celticminstrel is now known as celmin|sleep |
| 06:18 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 06:18 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 06:21 | <&McMartin> | Wow. 24 hours of silence in this channel. |
| 06:21 | | * McMartin is wrapping up a very long bumbershoot post that will be a Feat Of Literate Programming and possibly also a reason to not do that |
| 06:52 | <&McMartin> | This post is almost 64K long -_- |
| 06:52 | <&McMartin> | Maybe I should split it in two. |
| 06:55 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has joined #code |
| 06:55 | | mode/#code [+qo Vorntastic Vorntastic] by ChanServ |
| 07:03 | <&McMartin> | Oh that splits quite neatly |
| 07:03 | <&McMartin> | Okay then |
| 07:09 | <&McMartin> | https://bumbershootsoft.files.wordpress.com/2018/11/mac_cca.png |
| 07:24 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds] |
| 07:38 | <&McMartin> | https://bumbershootsoft.wordpress.com/2018/11/21/unfiltered-cocoa-powering-a-custom-widget/ |
| 07:41 | <&McMartin> | This is also my first attempt at explaining my way through Objective-C code in sufficient detail that someone used to C, C++, or Java can actually read it |
| 08:23 | | JustBob [justbob@Nightstar.Customer.Dissatisfaction.Administrator] has quit [Ping timeout: 121 seconds] |
| 08:24 | < simon_> | what's the best way to generate random sequences without repeats? Stackoverflow recommends generating a list and shuffling it. but what if my sequence is very long? |
| 08:25 | <~Vorntastic> | How long is the sequence you wish to generate |
| 08:25 | < simon_> | I remember this thing from years ago where if you pick a number that is coprime to the number of combinations and perform modular arithmetic, you'll go through the list one at a time in a non-sequential order. but you can still derive the order somehow. that doesn't matter strongly here, but it still feels a little silly. |
| 08:25 | < simon_> | it is 26*26*10*10*10 = 676000 |
| 08:26 | < simon_> | so not terrible. but still a lot considering a smart program would only be a few bytes. |
| 08:27 | <~Vorntastic> | Linear congruential generators are tried and true |
| 08:27 | < simon_> | so if you pick the coprime at random and pick one that is a little greater than the number of combinations, you'll go through the list backwards with a fixed distance. I wonder if there's a way to pick the coprime so that the order seems more "all around the place". |
| 08:27 | < simon_> | ah that's their name. |
| 08:28 | <~Vorntastic> | You multiply *and* add |
| 08:28 | < simon_> | ah! |
| 08:29 | <~Vorntastic> | You can even skip the add step if you're fine with not including the zero value |
| 08:32 | <~Vorntastic> | Well, the zero value and any non coprimes, which for that number is... Most numbers. |
| 08:39 | <&McMartin> | I'm also a fan of the xorshift family of PRNGs |
| 08:40 | <&McMartin> | https://en.wikipedia.org/wiki/Xorshift |
| 08:40 | <~Vorntastic> | Xorshift has a disadvantage in this case though |
| 08:40 | <&McMartin> | Oh wait |
| 08:40 | <&McMartin> | Random *sequence* |
| 08:40 | <~Vorntastic> | The modulus is always a power of 2 |
| 08:40 | <&McMartin> | Right, a random permutation is what they want here |
| 08:41 | <~Vorntastic> | Lcg with the right modulus is equivalent to an apparently random permutation |
| 08:44 | | * McMartin nods |
| 08:44 | <&McMartin> | I have had Bad Experiences with needing to use All The Bits on LCG |
| 08:45 | <~Vorntastic> | You can do Fisher Yates with... A lazy dictionary, which will grow for a while and then shrink |
| 08:46 | <~Vorntastic> | Gimme a minute I'll see what I get |
| 09:02 | < simon_> | hmm, a lazy dictionary. |
| 09:03 | <~Vorntastic> | Like, only store the swapped stuff; anything that hasn't moved is going to be completely predictable by program |
| 09:04 | < simon_> | so I realize that Python's 'random' is just a fancy combination of linear congruential generators. so https://docs.python.org/2.2/lib/module-random.html suggests (I know, 2.2, but the 3.x docs don't have this page) that I could use the internals of the 'Random' to generate a random permutation of a list in-place. |
| 09:05 | <~Vorntastic> | It is no longer |
| 09:05 | < simon_> | oh. |
| 09:05 | < simon_> | since this is Python, I figure that the interface I'm looking for is an iterable over a list. whether this is lazy via an LCG or a lazy dict could be an implementation detail. |
| 09:05 | <~Vorntastic> | Random.shuffle is a thing |
| 09:06 | < simon_> | but it returns a materialized, shuffled list, doesn't it? |
| 09:06 | < simon_> | wait, maybe it doesn't. |
| 09:06 | <~Vorntastic> | It does it in place |
| 09:07 | < simon_> | but I'd need the entire list that it does this over. since I'm doing it over an integer range, I should be able to do it without the whole list of integers in the range. |
| 09:07 | <~Vorntastic> | If you insist on an iteravle that isn't a reified list then you're doing something else and this lazy dictionary is a thing give me a minute |
| 09:10 | < simon_> | OK. :-) |
| 09:17 | <~Vorntastic> | https://docs.python.org/3/library/random.html |
| 09:18 | <~Vorntastic> | (btw) |
| 09:29 | <~Vorntastic> | Fuck me it got deleted |
| 09:30 | <~Vorntastic> | (doing this on my phone, so it's a little harder) |
| 09:40 | <~Vorntastic> | https://ideone.com/sqU2zb |
| 09:41 | <~Vorntastic> | Inner functions working the first time who accepted my black goat offering today |
| 09:42 | <~Vorntastic> | I am not certain of the size heuristics on swaps but a guess is like n/3 |
| 09:42 | <~Vorntastic> | Is the maximum size |
| 09:43 | <~Vorntastic> | Or because combinatorics is perverse like that, 1/e |
| 10:58 | <~Vorntastic> | Yeah okay I'll believe n/e |
| --- Log closed Wed Nov 21 11:45:21 2018 |
| --- Log opened Wed Nov 21 12:00:28 2018 |
| 12:00 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 12:00 | | Irssi: #code: Total of 31 nicks [24 ops, 0 halfops, 0 voices, 7 normal] |
| 12:00 | | mode/#code [+o TheWatcher] by ChanServ |
| 12:00 | | Irssi: Join to #code was synced in 16 secs |
| 12:03 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 12:03 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 12:22 | <@TheWatcher> | Oh, for fucks sakes |
| 12:22 | <@TheWatcher> | python manage.py runserver |
| 12:22 | <@TheWatcher> | Performing system checks... |
| 12:22 | <@TheWatcher> | Segmentation fault |
| 12:22 | <~Vorntastic> | Neat |
| 12:22 | <@TheWatcher> | Just what I fucking need |
| 12:23 | <@TheWatcher> | On top of my fucking phone line going flakey |
| 12:29 | | * TheWatcher learns of faulthandler |
| 12:32 | <@TheWatcher> | File "....blabla../env/lib/python3.6/site-packages/MySQLdb/connections.py", line 359 in set_character_set |
| 12:32 | <@TheWatcher> | sigh |
| 12:41 | <@TheWatcher> | FFS |
| 12:47 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has quit [Operation timed out] |
| 12:47 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has joined #code |
| 12:47 | | mode/#code [+ao Reiver Reiver] by ChanServ |
| 13:39 | | celmin|sleep is now known as celmin|away |
| 13:45 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [Connection closed] |
| 13:45 | | gnolam [lenin@Nightstar-4de.ala.4.155.IP] has joined #code |
| 13:46 | | mode/#code [+o gnolam] by ChanServ |
| --- Log closed Wed Nov 21 15:55:47 2018 |
| --- Log opened Wed Nov 21 15:55:53 2018 |
| 15:55 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 15:55 | | Irssi: #code: Total of 32 nicks [25 ops, 0 halfops, 0 voices, 7 normal] |
| 15:55 | | mode/#code [+o TheWatcher] by ChanServ |
| 15:56 | | Irssi: Join to #code was synced in 21 secs |
| 16:21 | <~Vornicus> | Huh. 254ish for piles of size 1000. |
| 16:33 | <~Vornicus> | simon_: so I did some checking, check this out: for size 1000 shuffles, the list would weigh in at like 4.5 kB; the dictionary used in my thing at like 10kb at its largest. |
| 16:43 | <~Vornicus> | for 6760000, I get... 30MB for list type and 100MB for lazy dictionary. I think we're looking at "not worth it" land, except that we're frontloading the processing on the first. Not that it's *that* hard to do shuffle anyway |
| 16:45 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has quit [[NS] Quit: Connection closed for inactivity] |
| 16:58 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code |
| 17:46 | <~Vornicus> | suppose you could rejigger so that it actually uses the list and just yields on each iteration |
| 17:59 | <~Vornicus> | https://ideone.com/HgP1Pk which is way shorter |
| 20:40 | | Reiv [NSkiwiirc@Nightstar-ih0uis.global-gateway.net.nz] has joined #code |
| 20:40 | | mode/#code [+o Reiv] by ChanServ |
| 20:44 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Operation timed out] |
| 21:52 | | Degi- [Degi@Nightstar-350dt7.dyn.telefonica.de] has joined #code |
| 23:10 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has quit [Ping timeout: 121 seconds] |
| 23:20 | | Degi- [Degi@Nightstar-350dt7.dyn.telefonica.de] has quit [Ping timeout: 121 seconds] |
| 23:23 | | Degi [Degi@Nightstar-350dt7.dyn.telefonica.de] has joined #code |
| 23:40 | | gnolam_ [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 23:40 | | gnolam [lenin@Nightstar-4de.ala.4.155.IP] has quit [NickServ (RECOVER command used by gnolam_)] |
| 23:40 | | gnolam_ is now known as gnolam |
| 23:40 | | mode/#code [+o gnolam] by ChanServ |
| 23:57 | | Degi [Degi@Nightstar-350dt7.dyn.telefonica.de] has quit [Ping timeout: 121 seconds] |
| 23:59 | | Degi [Degi@Nightstar-350dt7.dyn.telefonica.de] has joined #code |
| --- Log closed Thu Nov 22 00:00:26 2018 |