Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2015-02-07

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:04 pdurbin "It looks like you're trying to do bioinformatics in Excel" --Clippy https://twitter.com/philipdurbin/status/563715579375521793
04:04 jberger pdurbin++
04:04 jberger excel is always the wrong chioce
04:17 sivoais I recently saw some software that hooks Python into Excel
04:18 * jberger runs away screaming
04:19 sivoais <http://xlwings.org/>, <https://datanitro.com/>
04:19 sivoais yeah... if I was going to use Python, I'd never want to have Excel in the mix
04:20 sivoais Excel has data analysis tools, but they are all mediocre. The main thing it has going for it is Office integration.
04:21 jberger also, the whole concept is WAAYYYYY too close to what I was doing for bank of america
04:31 * sivoais owns a VBA 2003 book from Microsoft Press
04:31 sivoais it's bigger than most of my engineering textbooks
04:33 sivoais I think the only bigger one I have is Sedra & Smith "Microelectronic Circuits"
09:12 sivoais joined #pdl
11:58 pdurbin jberger: yeah, but what I like about https://github.com/datacarpentry/datacarpentry/blob/master/lessons/excel/ecology-examples/00-intro.md is that it meets scientists where they are.
11:58 pdurbin "In this lesson, we will assume that you are most likely using Excel as your primary spreadsheet program- there are others (gnumeric, Calc from OpenOffice), and their functionality is similar, but Excel seems to be the program most used by biologists and ecologists."
11:59 pdurbin which is why we're trying to support Excel in our app: Ingest: Investigate common uses of Excel for research data in order to define excel ingest requirements. · Issue #585 · IQSS/dataverse - https://github.com/IQSS/dataverse/issues/585
12:33 jberger excel can export csv, imo you should meet them there
12:36 pdurbin right. that's "4. Export to a text based format like CSV."
12:37 pdurbin but this may be non-obvious to biologists
12:37 pdurbin which is why I think that tutorial is helpful
12:39 pdurbin in time biologist might think, "Wait, why am I using Excel in the first place?" :)
12:45 jberger pdurbin: so a similar thing exists for physics, not excel but mathematica
12:46 jberger it's the crutch that usually works enough that they don't learn more
12:46 jberger but then when they need more they are left high and dry
12:46 jberger and I blame departments for getting lazy
12:47 jberger I did an informal poll in my department once and asked the professors where they learned to program
12:48 jberger nearly all said that they had had a required course somewhere along their studies
12:48 jberger now we didn't even offer one!
12:49 jberger most students get by with mathematica now, but I wonder if it's because they are not encouraged to look elsewhere, and without that, they find the easiest thing that works for them
12:50 jberger problem is, by the time you get to anything interesting, it isn't good enough anymore
12:51 jberger then you have to do like I did, learn a real language on your own out of desperation
12:52 jberger all the students that could program had either taught themselves or their prof had trashy them the basics at least
12:53 jberger now what kind of code style does that encourage
12:53 jberger ?
12:53 jberger they all have to discover floating point on their own
12:54 jberger tests and doc are unheard of
12:54 * jberger rants
12:55 jberger to summarize: I blame lazy departments
13:05 pdurbin heh. sure. makes sense
13:06 pdurbin part of my interest in this is my wife asking how she could avoid using proprietary tools like JMP
13:07 pdurbin her data wasn't even in Excel. It's in a table in a Word doc.
13:07 pdurbin so I extracted it into a csv
13:08 pdurbin the next step is to figure out some R to run the right model
13:08 pdurbin which a friend estimates will be 5 lines of R
13:18 pdurbin the model being Mantel-Cox: http://en.wikipedia.org/wiki/Log-rank_test
16:20 jberger pdurbin++
17:49 pdurbin :)
20:16 pdurbin mmmm, Smart::Comments - https://twitter.com/philipdurbin/status/564152937631744003 :)
21:18 sivoais joined #pdl
21:51 jberger the answer is that Smart::Comments is a source filter, so it can get the name easily
21:52 pdurbin yeah. and people seem to run screaming from source filters, probably for good reason :)
21:55 jberger normally that's true, but Smart::Comments is an exception in my mind
21:55 jberger it is explicitly not for production use
21:55 jberger it is a debugging aid
21:56 jberger and that way you don't pay for it if you don't need it
21:56 jberger I used to do that, but now I do the Mojo way of compile-time optimization of debug statements
21:57 jberger use constant DEBUG => $ENV{MY_DEBUG};
21:57 jberger then
21:57 jberger print "some debug" if DEBUG;
22:20 pdurbin right, you just comment out "use Smart::Comments" to disable it
23:47 jberger actually I have usually seen people not include that and do "perl -MSmart::Comments ..." to enable it

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary