Slow

>slow
>botnet
>start indexing at 1

seriously, why are people using this shit?

Other urls found in this thread:

youtube.com/watch?v=91RZYO1cv_o
twitter.com/NSFWRedditImage

indexing at 1 is confier desu fampai

youtube.com/watch?v=91RZYO1cv_o

forced to use it in the uni for lack of better alternatives

I had to use it for 3 semesters, it was hell.

even fucking FORTRAN is a better alternative

because most who use it are engineers or scientists who suck at programming, they just got used to it

too scared to learn python

People didn't have anything better.

Nowadays, I'd propose Jupyter, Zeppelin, Beaker etc. are usually better alternatives for computer programming literate people. Because you get to fight absolutely terrible syntax far less often, and because they integrate a lot of powerful tools.

The problem is that most people that learned Matlab never learned anything else... kinda like the crowd that always used MS Access/Excel.

What is the open source alternative of this?
My country ISP's have banned all torrent sites and tor is very slow. Can't pirate this.

You might want to check out Python with Spyder, Gnu Octave or Sage. There are literally 20 or more free options out there. But don't expect to get a nice documentation with lots of examples like Matlab, tho. Google is your friend.

GNU Octave

Octave is pretty compatible with Matlab. Matlab is so big there's just not enough people working for free to handle all the little domain specific add ons but Octave Forge does exist. Octave has a pdf manual that I found pretty good. The guys in #octave on Freenode are helpful as well.

Octave, but it sucks even more than matlab. Just learn how to use python and numpy and you are set for life

I'd say Jupyter, Zeppelin, Beaker, Scilab, Sage... for something nicer.

Another option, couldn't ever bring myself to like Octave's syntax or help or other things. It can do the job, but it always feels like it kinda RESISTS doing it.

muh matrix calculus

Use DDL sites.

Because the toolboxes it offers aren't present on the alternatives. The Fuzzy Logic one was really useful for me a few semesters back.

What, seriously? Python can be taught to 8-year-olds. It's the simplest, comfiest programming language there is.

One based indexing is the superior convention imo.
>inb4 someone posts that Dijkstra quote
If you actually read it you'll see that his whole argument for zero rather than one based indexing is that he thinks it looks better on an aesthetic level.

mathfags always index at 1

>tfw I had to use Maple for calc 1-3

I never asked anyone because I thought it might be a stupid question, but is there a good reason to index from 0?
It wouldn't change anything in memory if the language just started indexing at 1 rather than 0, would it? Assuming 1 replaced 0 anyway, not that 0 was skipped.

if you like starting at 1 better, then your retarded
for(int xiNigger=0; xiNigger

Matplotlib

Nah, it's unreadable implied shit

The reason is because they basically use a pointer and an offset

so just taking the 0 index is simply taking the pointer (pointer + 0)
taking the 1 index is taking (pointer + 1)

look up how C arrays can turn into pointers when passed as parameters

This.

>indexing from 1 is bad because i've already memory-mapped inexing from 0 and cannot learn otherwise

I'll read up on it. Thanks

also you can change to 1 indexing by subtracting 1 before applying it to the array but then you probably increase off by 1 errors which is probably why you only see it in higher level languages where it wont blow up

> slow
Depends for what. For most scientists, it's good enough. They don't know how to code anyways and courses are more readily available (read: sold to universities)

>botnet
Non-neckbeards don't even know what that is.

> start indexing at 1
Because people start counting at one. The language is designed around people, not machines.

But bottom line is that the software was built to support the science domain. There's a full vertical stack of services to teach and support users, while the language is simple enough (although fucking retarded in some areas) to be understood by simple folks. Another point is that the software is certified with regards to results accuracy. Finally, the various plugins serve specific domains (simulink for signal processing example) and there aren't real alternatives for those.

seriously learn python it will give you a better image in your CV and more skills in a general way
>u already know some bash and gnuplot

matsheit is wrong (have my pov so you are free to read or not)

I will graduate this year in CFD, and when i see my peers open matsheit to plot a stupid figure it a waist of time. The whole interface is clearly too much. Sure it will offer you to allocate list faster and some other tweaking but honestly if you want to build a code the diy philosophy remain the best. So you ll understand what you are doing.

It will be a botnet as long as source are not open. Honnestly, this code is a blackbox and for a scientific use it is always better to know how calculs are perform (In order to not say sheit). Plus if this software is as good as it pretend to be, they doesnt need to hide source.

Starting index at one ... srsly this aint a real argument.

For all its shortcomings, matlab is trying to tackle a different usage case than python. Sure, you can do anything in python that you can do in matlab if you're good enough at it, but matlab has a lot of built-in features to handle the type of math used in engineering and science. There's also a lot of code floating out there, written in matlab, for various engineering and scientific problems.

Personally, I think Wolfram Mathematica is better if you're doing math stuff, and most people who do math stuff get it for free through their university or institution.

Numpy and Scipy are the best Free as in Freedom™ alternative, if you're one of the hippy types into that sort of thing.

the whole indexing at 0 or 1 thing really just comes down to what you're doing.

If you are working with offsets of something like a memory address, use 0
If you are counting something, use 1

a language that uses both would be disgusting, that's why C uses 0 and languages like matlab and lua use 1

Fortran master race reporting in, I can just define my own indices whenever I define a matrix.

You want that shit to start at 0? Cool.
real A (0:100)

You want that shit to start at 1? Fine.
real A (1:101)

You want that shit to start at -23? Fuck me up senpai.
real A (-23:77)

>Just fuck me up with off by one errors