IRCLogs • ArmaNelgTron.tk

Searching from 2020-07-31 00:00:00 to 2020-07-31 23:59:59.999999.
Query completed in 0.50 seconds
[2020-07-31 03:28:52] <-- luke-jr has quit (Quit: ZNC - http://znc.sourceforge.net)
[2020-07-31 03:32:13] --> luke-jr has joined the channel
[2020-07-31 04:04:21] <-- danar has quit (Ping timeout: 256 seconds)
[2020-07-31 04:05:18] --> danar has joined the channel
[2020-07-31 07:36:41] <-- luke-jr has quit (Quit: ZNC - http://znc.sourceforge.net)
[2020-07-31 07:39:34] --> luke-jr has joined the channel
[2020-07-31 12:30:20] <Z-Man> They don't make that easy.
[2020-07-31 12:31:20] <Z-Man> Also, launchpad wants plain text, Steam bbcode, and itch, sourceforge and download.armagetronad.org take markdown. The main site is plain html.
[2020-07-31 12:33:15] <Z-Man> And at least this time, the text needed to be slightly different for each. Permanently, too, probably. I probably should not write about serverside improvements on platforms where we don't distribute the server at all (itch.io) or where it is well hidden (steam).
[2020-07-31 19:24:35] <Lucifer_arma> That's a job for txt2tags
[2020-07-31 19:25:11] <Lucifer_arma> more specifically, it's a job for the main site, and my new cms for it that's not ready to take it over
[2020-07-31 19:25:39] <Lucifer_arma> but when it is, it should be reasonable to have a module that'll export news items to each of the needed formats, with substitutions for text that needs to be slightly different, etc.
[2020-07-31 19:25:52] <Lucifer_arma> now then, my newest problem to think about:
[2020-07-31 19:25:57] <Lucifer_arma> how do you know when a stock price changes?
[2020-07-31 19:26:48] <Lucifer_arma> you can't just do a != comparison, because a $5 stock is still a $5 stock if the price is $4.50 or $6, but averages around $5
[2020-07-31 19:27:13] <Lucifer_arma> but if you say it has to have a certain percentage difference, say 5%, then you get a drift
[2020-07-31 19:27:52] <Lucifer_arma> so your $1 stock could go up to 1.05, be considered the same price, and then it goes up to 1.10, which is within 5% of 1.05, meaning it hasn't changed
[2020-07-31 19:28:50] <Lucifer_arma> then it goes up to 1.15, then to 1.20, then 1.26, and you see what's happening.  It's actually increasing, but because each value is 5% or less higher than the previous value, you've missed the increase
[2020-07-31 19:29:27] <Lucifer_arma> I need to do a basic analysis of a stock's history and find the minima/maxima, which means being able to tell when the price changes direction
[2020-07-31 19:30:16] <Lucifer_arma> I was thinking of saying a stock price is an object, and making a class for it.  Then I'd feed in each new value, and if the value is within 5% of the average, it's considered the same price and appended to a list inside the object
[2020-07-31 19:31:00] <Lucifer_arma> that list is the one that's averaged together to determine if the price has changed.  Once I try to give it a price that's outside 5% of the average, then the price would be "changed", and the average would then become the datapoint
[2020-07-31 19:31:22] <Lucifer_arma> and that datapoint would be placed in the middle of the date range for all the prices
[2020-07-31 19:32:30] <Lucifer_arma> then a maximum would be found if the stock price was previously increasing, then it stayed the same for a bit, then it started decreasing.  The average of the price where it stayed the same would be the maximum, and the date on which it occurred would be the date that's smack in the middle of the date range of all the prices used to determine that maximum
[2020-07-31 19:33:21] <Lucifer_arma> I like this idea, but it seems almost as naive as simply comparing today's price to yesterday's to see if it changed.  I always feel like using a simple average is naive, even when it's not.
[2020-07-31 20:46:20] <Lucifer_arma> ok, after doing the breakdown on social media controlling the flow of information and how their algorithms were exploited in the 2016 election and noticing the microsoft isn't one of them,
[2020-07-31 20:46:31] <Lucifer_arma> I'm starting to consider that microsoft may have jumped sides
[2020-07-31 21:40:43] --> Z-Man- has joined the channel
[2020-07-31 21:40:43] -!- Z-Man changed nick to Guest85246
[2020-07-31 21:40:43] <-- Guest85246 has quit (Killed (verne.freenode.net (Nickname regained by services)))
[2020-07-31 21:40:43] -!- Z-Man- changed nick to Z-Man
[2020-07-31 22:24:45] <ct|kyle> Lucifer_arma: what are you trying to do with stocks?
[2020-07-31 22:26:14] <ct|kyle> Lucifer_arma: there's a lot more to it then just the price, definatly works best with some sort of historical comparison, maybe like 30 day moving avarage or something
[2020-07-31 23:56:20] <Lucifer_arma> ct|kyle: much more complex stuff than that
[2020-07-31 23:57:03] <Lucifer_arma> ct|kyle: but otherwise, trying to write a program that'll give me undervalued stocks, or find stock prices that are fluctuating in a predictable pattern (like cyclical stocks) so I can buy low and sell high
[2020-07-31 23:57:54] <Lucifer_arma> I'm imagining a table that stores a 7day and 30day moving average, at least, and I thought I could add a few fields to indicate how the stock's moving daily.  The four fields I have to work with for that are open, close, low, and high
[2020-07-31 23:58:30] <Lucifer_arma> rather than try to average those together (silly), I figured I'd just analyze each one.  So that'll be my most basic report on a stock's history.
[2020-07-31 23:59:00] <Lucifer_arma> I've just about finished the history retrieval stuff.  There's tuning to be done, and performance stuff, but it's doing the job correctly and not making duplicate records
View entire month
DISCLAIMER: These logs of public chat may contain some content which may not be appropriate for all audiences. Use at your own risk.
Logs from 2006-2009 pulled from wrtlprnft
Format changes at: 2015-08-25, 2017-02-20, and 2020-03-23. Times (2015 and later) should be Eastern.
. [About this site] [Credits]