Mathics – A free, light-weight alternative to Mathematica

https://mathics.github.io/

by dvfjsdhgfv • 7 years ago

360 83 7 years ago

mathics.github.io

O_H_E• 7 years ago

I would like to take this amazing opportunity to tell anyone who doesn't know about WolframAlpha to go give it a shot.

Shortest, simplest description I can give is that it is the best AI-less science-heavy digital brain out there. Or as Wikipedia puts it "a computational knowledge engine".

https://wolframalpha.com

Edit: apparently Siri and Alexa sometimes query WolframAlpha.

wjsetzer2• 7 years ago

While Wolfram alpha got me through Cal 2, why are you plugging it here?

slavik81• 7 years ago

WolframAlpha includes a lot of Mathematica features and many uses are free-as-in-beer. For example,it can evaluate Mathematica expressions like Series[x - n Sin[x], {x,0,5}]. http://m.wolframalpha.com/input/?i=Series%5Bx+-+n+Sin%5Bx%5D...

starpilot• 7 years ago

There's this search engine called Google too, it's really accurate.

zwegner4• 7 years ago

And I'd like to plug my favorite twitter account, Wolfram|Alpha Can't: https://mobile.twitter.com/wacnt/

It's a collection of queries that an AI could answer in theory, but are (probably) far beyond current technology. The questions are generally just super interesting.

Here's a few recent samples: "derivative of the shore length of Tasmania with respect to ocean depth" "total area enclosed by whole notes in a typical printing of the Brandenburg Concertos" "total duration of watch adjustments made by people changing time zones per day, if everyone had a watch and adjusted it" "volume of ink saved by the Times per year if they switched to American spelling"

harshreality2• 7 years ago

I think those questions are just an attempt to be silly.

Most of those questions are poorly specified, if you think about them, or they're trick questions with trivial answers (like zero) in some cases.

Then there's the occasional question that could be answered, not just by some future AI, but even current WolframAlpha if it had the right dataset. Like this:

minimum distance between a place with a 2-letter name and Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch

Thought it's unclear if the intent is great-circle distance or 3d distance.

If there's any doubt that plenty of these are intended to be silly questions...

focal length of the Eye of Sauron

chongli• 7 years ago

derivative of the shore length of Tasmania

That's a funny joke but it's not something for which any AI could provide a concrete answer. This is due to the fractal nature of coastlines and described as the coastline paradox [1].

[1] https://en.wikipedia.org/wiki/Coastline_paradox

searke2• 7 years ago

I'm someone who occasionally works on Wolfram|Alpha and also a big fan of the account. Glad to see you're still going. Some of your examples have made me think really hard about ontological and organizational problems I hadn't yet considered.

edjroot• 7 years ago

You can also get a few extra features on the mobile app, such as step-by-step solutions, if you pay a very small, one-time value (not sure how much).

ErotemeObelus5• 7 years ago

Be prepared to have Stephen Wolfram pull all the stops in order to destroy your project. He didn't get rich by letting free competition flourish you know.

EDIT: I downloaded the source just in case the author receives an offer he can't refuse.

techgipper3• 7 years ago

The developer works at Wolfram, just FYI.

https://medium.com/@poeschko

ErotemeObelus5• 7 years ago

He should double-check his contract to see if there's an "anything you do either inside or outside of Wolfram Inc. is automatically our property" clause.

Stephen Wolfram knows about this. He just hasn't taken action yet.

carlob• 7 years ago

1. This project has existed for quite some time and nobody has sued them

2. The originator of this (Jan Poetschko) has worked at Wolfram for several years now

3. I think at least a number of people at Wolfram understand that Octave has been an asset (or gateway drug if you prefer) for Matlab, rather than a threat

ErotemeObelus5• 7 years ago

I never said sue. I said pull all the stops to make the project and the developer go away (or coopt it). Maybe that includes suing but probably not given how expensive lawsuits are.

Usually companies are intelligent enough to use tactics that has the greatest likelihood of success: mafioso tactics. The developer won't, in all likelihood receive a lawsuit. He'll be given an offer he can't refuse.

hopscotch• 7 years ago

Wolfram don't make matlab. They make Mathematica.

williamstein• 7 years ago

Maybe Wolfram already effectively killed this project. It was actively developed for about 5 years (see https://github.com/mathics/Mathics/graphs/contributors), then Wolfram Inc. hired the lead developer 2 years ago, and basically all worked stopped...

ErotemeObelus5• 7 years ago

It appears as if my warning bore out then.

It's a good thing they didn't pull the plug on the github page. Now it's possible for someone to carry the torch.

marchairthiel• 7 years ago

Maybe Wolfram paid him $1 billion for the project...the possibilities are endless...

delhanty3• 7 years ago

I'm confused.

The lead developer up until 2 years ago, Angus Griffith's LinkedIn says he's at Google. [0]

In April 2016, Angus replied to a comment of mine "Mathics is still maintained. There are currently two of us working on it. The last release was v0.9 in March this year." [1].

The original creator Jan Pöschko's GitHub profile states "JavaScripter making the notebook interface of wolframcloud.com" [2]. His bulk contributions to Mathics stopped in 2012, but he still makes sporadic contributions to Mathics - see here he merged a pull request March 2019 [3] - suggests Wolfram tolerate the existence of Mathics because it's not a serious threat.

As the creator of Sage, what's your opinion of the legal risks of creating a Mathematica syntax compatible clone like Mathics in the light of Wolfram Research's fairly aggressive assertion of its IP "rights"?

[0] https://www.linkedin.com/in/angus-griffith-093433b0/

[1] https://news.ycombinator.com/item?id=11465064

[2] https://github.com/poeschko

[3] https://github.com/mathics/Mathics/commits?author=poeschko&s...

matt20192• 7 years ago

Tangential, but reminds me of Stephen Wolfram's story of how it got the name Mathematica. It was actually Steve Jobs's idea:

"Over the months after our first meeting, I had all sorts of interactions with Steve about Mathematica. Actually, it wasn’t yet called Mathematica then … For a while he wouldn’t suggest anything. But then one day he said to me: 'You should call it Mathematica.' … he told me his theory for a name was to start from the generic term for something, then romanticize it." (Wolfram, Idea Makers)

szemet2• 7 years ago

then romanticize it

You mean romanize? And not just on the surface (latin alphabet), but even using the actual word from latin (which is already romanized greek I guess) and of course all the related living languages...

edit: I always felt strange naming a software exactly the same name (in many language) as the field it operates in. Like naming a music creation software Music, or a biology related software Biology, etc. It is like equates the software with the whole field, suggesting that there is no difference between the software and its disclipine. It is a marketing trick that have a bit of bragging side, but of course it is common pattern, e.g: https://www.amazon.co.uk/I-Am-Football-Zlatan-Ibrahimovic/dp...

pdabbadabba2• 7 years ago

I'm not sure what you're getting it. I think GP meant "romanticize," i.e., "deal with or describe in an idealized or unrealistic fashion; make (something) seem better or more appealing than it really is."

(Though in this case the intended meaning is probably a little closer to "build mystique surrounding.")

I agree with your point about naming a product to cover the whole field. It does seem off-putting--although I think examples of it are pretty rare. (In fact, I can't think of any off hand...) In this case, "Mathematica" is far enough from anyone's ordinary English usage that it doesn't fall into that category for me. I don't believe I've ever heard anyone say the word "Mathematica" to refer to anything other than the product. It would be a different story, of course, if it were just called "Math."

weinzierl• 7 years ago

When the Swiss designer Max Miedinger created a new typeface in the late fifties, his client Stempel AG wanted to call it Helvetia, which is another name for Switzerland. Max Miedinger found it silly to name a typeface after a country and suggested Helvetica - the Swiss. That's what we've got Helvetica.

Gaelan• 7 years ago

I always felt strange naming a software exactly the same name (in many language) as the field it operates in. Like naming a music creation software Music, or a biology related software Biology, etc.

IIRC Microsoft did this (Word, Windows) with hope that the press would feel the need to use “Microsoft” to disambiguate, improving brand recognition.

jweir• 7 years ago

Philosophiæ Naturalis Principia Mathematica - Isaac Newton

https://en.wikipedia.org/wiki/Philosophi%C3%A6_Naturalis_Pri...

matt20192• 7 years ago

While I have this book in front of me: "all sorts of people were telling me that I needed to put quotes on the back cover of the book [A New Kind of Science] ... Steve said, 'Isaac Newton didn't have back-cover quotes; why do you want them?'"

notfromhere• 7 years ago

I definitely believe that he had nothing to do with naming it.

rolleTx• 7 years ago

If it is light-weight why don't they make web-app like Wolfram Alpha. I want to see an alternative to that.

thearn4• 7 years ago

SymPy Gamma (https://gamma.sympy.org/input/?i=integrate+x%5E2) sort of fits as an alternative, though it's not very good at parsing general questions/expressions.

codetrotter• 7 years ago

Woah! That is very cool indeed. Time to start using SymPy Gamma instead of Wolfram|Alpha. Surprising that their GitHub repo has so few stars. Only 400 stars at the time of this writing. I mean, 400 stars is respectable of course, but seeing how cool this is I think they ought to have a lot more stars.

corysama• 7 years ago

In big, bold text in the beginning of the description is the header for "Use Mathics online"

fireattack• 7 years ago

It goes to a website with "NET::ERR_CERT_DATE_INVALID"

TheRealPomax• 7 years ago

Mathematica and Wolfram Alpha are two completely different things, though. The latter is nothing even close to just a web UI for the former.

wjsetzer2• 7 years ago

I haven't had a lot of time to dig through the docs yet (also, I think your certificate is expired, btw), but this seems like a great alternative to Mathematica and the like.

This is definitely a project I can get behind.

groovybits• 7 years ago

Can confirm, the certificate for angusgriffith.com is expired.

sn6uv5• 7 years ago

Just renewed it (not that it'll survive the HN hug of death). It's been a few years since I've touched that website.

jhbadger• 7 years ago

This is a cute little project that can be deployed anywhere with pip install mathics -- I even have it on my android phone (under Termux)! It can't really compete with either Mathematica or Sage, but both are multi-gigabyte programs. Unfortunately the Jupyter kernel IMathics doesn't seem to work with the current release, although it does come with its own notebook support.

abc_lisper2• 7 years ago

Cert expired.

sn6uv5• 7 years ago

Just renewed it

abc_lisper2• 7 years ago

Thanks!

abawany• 7 years ago

Somewhat unrelated but I would like to also mention Maxima (http://maxima.sourceforge.net/), which used to be maintained by the late Dr. Schelter and is still updated. Here is a link to an old comparison between it and Mathematica: https://luckytoilet.wordpress.com/2014/08/11/beginners-compa... .

gnufx2• 7 years ago

It could be related though Fateman's latest Mockmma, linked from https://people.eecs.berkeley.edu/~fateman/. (He got a legal threat from Wolfram over the original version long ago.)

Other free general symbolic algebra systems are available, of course, like the venerable Reduce and Axiom.

enriquto• 7 years ago

I used to use mockmma a lot as an undergraduate, because it started much faster than mathematica and it could run exactly the same programs. I wonder why mathics does not keep exact compatibility of the language.

pfortuny• 7 years ago

Fateman was (is?) one of the main contributors!to maxima. Quite a helpful guy, also.

xrd• 7 years ago

Came here hoping for a discussion of the project itself but this is all information about Mathematica or the fact that the developer who started it now works there.

Does anyone have an opinion on the software itself linked here?

dvfjsdhgfv2• 7 years ago

I'm going through the exampleas and apart from some minor issues it looks like it's doing what it's supposed to do.

brian_herman__• 7 years ago

Author of the program uses let’s encrypt and the ssl certificate is expired for the documentation... https://mathics.angusgriffith.com/doc/

sn6uv5• 7 years ago

Just renewed it

dang• 7 years ago

A discussion from 2017: https://news.ycombinator.com/item?id=11459186

godelski2• 7 years ago

This needs examples front and center.

Seriously, why is this not more common? When I go to your landing page I want to see how awesome your thing is. A few examples is all it takes.

Show me something like this

In[6]:= Derivative[1][Sin]

Out[6]= Cos[#1]&

In[7]:= Derivative[2][Sin]

Out[7]=-Sin[#1]&

Don't make me go to documentation. Also don't compare it to Mathematica if it isn't as intuitive as Mathematica (seriously, it is wildly popular because simple things you don't have to look up. It works like you expect. There was a lot of planning to make it this way. People note the "Zen" of Mathematica). Here's an example of where Mathics is non-intuitive (besides if you remove the [1] from the above example)

n[13]:= x = 5

Out[13]= 5

In[14]:= Derivative[1][x^2]

Out[14]= 25'

In[15]:= f[x] := x^2

In[16]:= Derivative[1][f]

Out[16]= 2 #1&

In[17]:= f'[x]

Out[17]= 10

I know no system is perfect, but try to sell me on your idea before I go through your documentation and get frustrated that it doesn't work like the thing you compare it to.

sn6uv5• 7 years ago

Thanks for the feedback, the web UI does have a gallery link [0] with a few interactive examples [1]. I agree though, a screenshot like this would be a good addition to the homepage.

[0] https://pasteboard.co/IdIaJNo.png [1] https://pasteboard.co/IdIc7dc.png

godelski2• 7 years ago

I don't think that first image helps but that second image is great. That gets me excited and makes it look intuitive. Seriously, put that on the landing page.

Edit: Also that first example (from [1]) doesn't work for me

In[1]:= 1 + 2 - x * 3 x / y

Out[1]= 3 - 3 x ^ 2 / y

Is this a difference in the cli? I also notice I can't get any graphics there. No connection to matplotlib?

In[3]:= Plot[{Sin[x], Cos[x], Tan[x]}, {x, -3Pi, 3Pi}]

Out[3]= -Graphics-

dvfjsdhgfv2• 7 years ago

Thanks a lot for your great work! I installed mathics locally and the translate/sierpinski examples from the gallery don't seem to work. Do you have an idea why?

syockit• 7 years ago

Sometimes I confuse the meaning of light-weight to be something that "does simple thing quickly" instead of "easy to setup". The last time I played with Mathics (maybe 2 years ago) it was still much, much slower than Sympy.

axelclk• 7 years ago

The split between SymPy functions and Mathics wrappers around these functions, together with the pattern matching is probably responsible for the slower system. Therefore in the Symja project[0] we try to implement it all together in one (Java) library. Mathematica syntax can be tested in the MMA Console[1], but the documentation of the functions [2] is in the "normally used Symja syntax", which is much easier to use for the "common calculator user".

[0] https://github.com/axkr/symja_android_library [1] https://github.com/axkr/symja_android_library/wiki/MMA-conso... [2] https://github.com/axkr/symja_android_library/tree/master/sy...

Mathics – A free, light-weight alternative to Mathematica

Related Stories