00:26 Manifest0 left 01:24 hulk joined 01:25 kylese left 01:34 jmcgnh left 01:45 jmcgnh joined 01:59 jpn joined 02:04 jpn left 02:15 hulk left, kylese joined 03:17 jpn joined 03:21 jpn left 04:18 jpn joined 04:23 jpn left 05:48 jpn joined 05:52 jpn left 08:03 holyghost joined 08:12 sena_kun joined 08:16 holyghost left, holyghost joined, holyghost left 08:17 holyghost joined 08:18 holyghost left 09:18 Manifest0 joined 09:51 holyghost joined 09:52 holyghost left 10:07 holyghost joined 10:14 holyghost left 10:15 jpn joined 10:20 jpn left 10:24 jpn joined 10:29 jpn left 11:13 evalable6 left 11:16 gordonfish left 11:18 evalable6 joined 11:24 gordonfish joined, synthmeat left 11:27 synthmeat joined 11:37 jpn joined 11:49 jpn left 11:54 sena_kun left 12:17 jpn joined 12:22 jpn left
tbrowder anyone have a use case for a CSV file with row names but no header? i'll chk back later afk& 12:43
lizmat in my experience, CSV files come with or without headers 12:44
13:34 jpn joined 13:39 jpn left 13:58 Manifest0 left 14:06 jpn joined 14:11 jpn left
ab5tract ooc, how would you do row names without a header? 14:16
greenfork What does "iron-manning" mean in the context of this article, under the "Archer" heading? strangelyconsistent.org/blog/perl-6-is-my-mmorpg 14:21
antononcube @ab5tract "how would you do row names without a header?" -- Just using integers corresponding to the columns? (This what I do in my some of "Data::*" packages.) 14:27
ab5tract ah, I was confused. I misread that tbrowder was referring to column names. 14:29
14:33 jpn joined
ab5tract and of course you can map column ids to column names on the consumer side. but I don't see how the parser would interact with that at all beyond taking a list of column names as an argument to use as a lookup 14:34
14:40 jpn left 14:47 [Coke] left, [Coke] joined
tbrowder back again. well, my plan is to have my csv handler do some fancy tricks. and then you can do things like keep a table of monthly (columns) recurring expenses (rows) and summarize totals and averages for the expenses by name 15:08
with column names you can get total expenses for, say, Jun 15:29
15:36 jpn joined 15:41 jpn left 15:48 bdju left 15:51 bdju joined 15:53 jpn joined 15:58 MasterDuke left, jpn left 15:59 jpn joined
jdv ab5tract: yeah, i guess its harder 16:04
16:05 jpn left 16:10 abraxxa-home joined 16:29 haxxelotto joined 16:37 jpn joined 16:40 haxxelotto left 16:42 jpn left 17:13 Manifest0 joined
tbrowder ab5tract. my project is a CSV file handler which include data management after parsing the file. 17:30
17:37 sena_kun joined 17:38 jpn joined 17:43 jpn left
ab5tract Sure. I guess I don’t understand the question about row names without a header. 17:52
Specifying column names with a list of names passed to the parser is definitely important for a CSV library. IMO 17:56
Apologies if I‘m still misunderstanding 17:57
librasteve tbrowder: to answer your question about row and col - here's my deck on raku Dan from tprc 2022 static.sched.com/hosted_files/tprc...Dan-v2.pdf ... scroll down through the BS until you get to the DataFrame Anatomy slide and you can see the "top level design" 18:06
in a nutshell a DataFrame (ie a spreadsheet) looks like this 18:11
role DataFrame { has Any @.data =[]; #2d [m;n] has Int %.index; #row index has Int %.columns; #col index }
and you access like this 18:12
say df[0;0]; df[0;0] = 3; # set value # Cascading Accessors (ok to mix Positional and Associative) say df[0][0]; say df[0]<A>; 18:14
ie there is a row index and a col index - both are hashes that you can address by row name and col name
DataFrames as popularized by Pandas and so on are good for data science where you process big data in columnar format ... so they are a bit overkill for a spreadsheet 18:16
18:18 jpn joined
i am not suggesting you do cascading accessors (although imo it is cool to go say df<x><y>; and so on) there are certainly less painful ways to use hash keys to address cells ;-) ... and Dan does not handle multiple header rows 18:19
fwiw I can see the benefit of a less formal / more lightweight version of spreadsheets in raku ... when you have a v1 I can try and work out how to easily go to/from Dan and also would be cool to make it work with my other Net::Google::Sheets module 18:21
18:37 jpn left
tbrowder librasteve: gosh, my bad, i now remember seeing the slides and thinking: this is a seriously powerful thing for an egineer or scientist!! but i think i dismissed it 'cause i couldn't see it fitting into my little world view then. 18:52
my current thing started because i needed a good csv handler for my calendar event data input...and you probably know the rest...just add one more method... 18:55
19:12 jpn joined 19:19 MasterDuke joined 19:23 abraxxa-home left 19:24 jpn left 19:33 Donjon joined, Donjon left 19:36 haxxelotto joined
tbrowder librasteve: can i ask about yr use of the raku OAuth2? i 20:51
teatime2000 yeah ofc they still happen, I never said there were none 20:56
21:04 sena_kun left 21:05 sena_kun joined
librasteve tbrowder: sure - it works well for me ... but for Google APIs there is the console app API key setup of course ... all that is explainedin the OAuth2 module 21:13
if you want to use it, why not take a look at class Session {...} in Net::Google::Sheets ... step one would be for you to take it for a spin and I'm happy to help if needed 21:14
ab5tract teatime2000: I definitely didn't mean to imply that you said that, I was trying to clarify that _I_ wasn't saying that 21:20
antononcube weekly: rakuforprediction.wordpress.com/20...-exercism/ 22:09
notable6 antononcube, Noted! (weekly)
22:15 haxxelotto left 22:16 haxxelotto joined 22:30 haxxelotto left
tbrowder librasteve: thnx 23:11
23:16 bdju left 23:18 bdju joined 23:29 sena_kun left 23:59 Manifest0 left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!