« first day (1660 days earlier)      last day (3122 days later) » 

halirutan
halirutan
00:00
@psimeson OK. Btw, don't copy so much data (and even a lot of code) here. Use for instance pastebin and just provide the link.
psimeson
psimeson
@halirutan ok
halirutan
halirutan
@psimeson Okey.. I think I'm going to apply the correct amount of ego-shooter now before finally going to bed.
psimeson
psimeson
@halirutan okay
 
1 hour later…
JasonB
JasonB
01:21
Anybody else here sometimes feel like Dana Carvey's Grumpy Old Man? Reading this post I keep thinking "In my day we wrote our own plotting functions and it was hard! We had to combine Graphics primitives into a cohesive, publication-worthy figure and we liked it!
Just in case y'all are too young for that reference, youtu.be/BbU4Cb4A4-o
Michael E2
Michael E2
02:15
@JasonB Somewhere I have plotting libraries I wrote in PostScript, 2D & 3D (line drawings only). I wrote the initial piecewise-plotter to replace part of it. Mainly for class materials. There used to be a TeX system you could write Postscript directly in the input file.
J. M.
J. M.
02:36
@JasonB The kids are getting spoiled with each new version. :)
 
3 hours later…
psimeson
psimeson
05:37
@halirutan Do you think Union[data,SameTest->(First[#1] ==First[#2]&)] is better way to get rid of duplicates in my problem?
kirma
kirma
06:08
$Version (* "11.0.0 for Linux x86 (64-bit) (July 28, 2016)" *)
On Wolfram Programming Cloud, or whatever they call it.
 
4 hours later…
halirutan
halirutan
10:10
@psimeson No, because Union sorts the data additionally. We had a long post about why DeleteDuplicates is faster.
 
1 hour later…
halirutan
halirutan
11:23
@psimeson Look here or here.
 
1 hour later…
user3183724
user3183724
12:35
Counter intuitive question, but is there a dashed variant of joined for listplot? As in, I have a bunch of points that form a curve, but I'd rather they form a dashed line (and joined would be a solid line)
Sascha
Sascha
Try PlotStyle->Dashed
user3183724
user3183724
Hm that just gives the normal dots from ListPlot
Sascha
Sascha
PlotStyle and Joined are two independent settings
you need to keep Joined
Example: ListPlot[Prime[Range[25]], Joined -> True, PlotStyle -> Dashed]
user3183724
user3183724
Wow, yeah. Amazing that you can deduce the stupid mistake I made with so little words :) That does it!
halirutan
halirutan
12:49
@user3183724 @Sascha And btw, ListLinePlot was invented so that you don't have to give the Joined->True option because it seems many people like to connect their points with lines.
user3183724
user3183724
@halirutan Interesting! Seems a bit unnecessary..
halirutan
halirutan
@user3183724 Well, it's there for a long time already and you have to type less as it is shorter to write 
ListLinePlot[Prime[Range[25]], PlotStyle -> Dashed]
:)
xzczd
xzczd
@kirma The document is also available now: reference.wolfram.com/language/guide/…
reference.wolfram.com/language/guide/…
 
3 hours later…
Meta
Meta
16:03
1
Q: totally confused about login: is it for stackexchnage or just individual site, e.g., mathematica.stackexchange

murrayI had or have at least two accounts, one originally with mathematica.stackexchange.com for which my login was via Facebook (or possibly Google), which in turn used, let's call it, email-address-1. Subsequently I went also to tex.stackexchnage.com, got a "Join this community" link when I tried to ...

support user-accounts login
Searke
Searke
@JasonB I recently have been suggesting to friends of mine that kids need to be taught how to use slide rules. People who grew up using slide rules inherently feel that precision and accuracy affect calculations in a way that people who grew up on calculators do not.
JasonB
JasonB
16:44
@Searke - I never had the pleasure myself, just followed recipes in chem lab for how to report error at the end of a long calculation involving measurement results, don't really have an intuition for it though
J. M.
J. M.
My personal belief is that people who are in the business of computing should be made to read Numerical Methods that Work before they get themselves into deeper trouble.
@JasonB Back when I still did lab work, our SOP was always to keep as many digits as possible, and only round the final numbers for the reports.
Searke
Searke
@J.M. It'll never happen. I could also say the same about software development skills like design patterns, which the scientific computing community woefully lacks.
J. M.
J. M.
@Searke Inertia is a hard thing to push back against. ;)
Searke
Searke
they're two very different cultures. It's hard to convince someone with a Phd in numerical simulations of black holes that they have a gap in their programming knowledge which would be filled by an undergrad course.
J. M.
J. M.
Might as well: I've found in a lot of cases that the higher the degree, the worse the code organization gets. That is: I've seen undergrads write more readable code than Ph.D.'s. I still don't know why.
Searke
Searke
16:54
@J.M. Easily. I want to chalk it up to hubris.
J. M.
J. M.
Yes, hubris is a major factor. Also, maybe the usual problem of things being clear and obvious to them not necessarily being clear and obvious to other people.
which is why I cherish papers where the code is actually understandable in two or three glances.
William
William
^ I like papers with code to begin with. I have read some that I don't think they want their experiments repeated
J. M.
J. M.
17:15
@William Some papers give that impression, yes. Any vaguer, and they'd be patent applications.
psimeson
psimeson
17:28
@halirutan I also found that DeleteDuplicatesBy is much faster than Union but Union seems to do much better job at getting rid of Duplicates than DeleteDuplicatesBy
JasonB
JasonB
@J.M. When I was doing my PhD and for a while after, I was the only one who ever used my code, so readability wasn't all that important to me. Efficiency was only an issue occasionally, usually I would just go for something that worked and then move on
Searke
Searke
17:50
@JasonB Software engineering and Zen buddhism are essentially the same thing. They both emphasize that the self doesn't actually exist. So you aren't allowed to make that excuse.
I made the mistake of thinking I was myself so many times. Then I'd look back at my own code in horror as if it were written by some monster.
Even when you write code for yourself, you have to write code as if you are writing for someone else.
J. M.
J. M.
When I write code, one of my underlying assumptions is that my future self would be a rather different person. So, that squares with my wanting to keep my code as maintainable as possible.
Searke
Searke
My future-self has consistently been a horrifically negligent idiot for as long as I've been programming.

« first day (1660 days earlier)      last day (3122 days later) »