Post your best R/mathematica/matlab/python graphs

i like graphs, I like to use them to understand concepts that I'm dealing with in mathematics, it helps visualize some things

could I use python to view composite functions and other non-basic stuff? you know a good library that can do that? It would be great if there were libraries that let you graph integrals, for example. I mean, anything more advanced than the usual math shit that comes in programming language standard libraries (abs, trig, etc).

you reek of autism

I want to learn classical mechanics, optics and re-learn linear algebra and calculus for a 3d computer graphics course, I want to have a good tool for plotting functions and data that I can use to understand some shit more easily by looking at visual patterns

>could I use python to view composite functions and other non-basic stuff? you know a good library that can do that?

numpy and mathplotlib should do it I guess

thank you very much, I'll go with python

for that pure mathematical purpose without a doubt mathematica is the best. It is the most equip at visualizations outside a statistical framework.

example: raspberrypi.org/blog/give-us-your-best-mathematica-one-liner/

quora.com/What-is-the-coolest-thing-you-can-do-with-Mathematica

Graphing composite functions would be best on Matlab/Mathematica (both are equal in this case). Python would be fine but you'd have to find those libraries on your own since I don't know them. Based on what I had to do, numpy and matplotlib worked fine.

Also desmos.com/calculator is useful as well for easier functions.

Based on that, I'd recommend either pirating Matlab (2015 version) or going with python.

without a doubt R:ggplot2 is the superior statistical/data analysis visualization tool, with python+numpy+matplotlib a close 2nd

...

...

I see. Looks like that python combination of libraries is mentioned a lot, so I should learn it. I'll avoid R just because python is a more valuable skill than knowing R since it applies to more types of jobs, so it increases employability more IMO.

I wanted to avoid matlab / mathematica but I guess they're worth a pirate in case python won't do the job

thanks, yeah, I've seen lots of interesting shit being done with it, it's fun staring at those graphs and trying to understand the patterns that form in them even if you don't fully understand the mathematical processes that cause those graphs to be the way they are

yea python is also more robust than R, in my industry (quant finance) python is the main tool and R is a side tool. helps to have but wont make or break you

continuing to chart post ggplot2 stuff

this is the best one i think i have ever seen fromr

If you know Python, learning R is piss-easy, and Python just still can't compete with R for basic capabilities for data manipulation and exploration.

R is so easy even psychologists use it.

...

>no title

how do you get the colors to blend on the overlaps like that in ggplot?

contributing also. don't have much on this machine.

...

...

...

>tfw bullshitted way through scientific computing course
>got 100% on all the coursework

Protip: figures that get published often have no titles at all. That's what legends are for.

...

They're sooo preeeeetty

Possibly this. Made in R.

Made in R

Yes. My dick is so hard. So fucking hard.
>I want to fuck that trebuchet

Give me a 2D shape and I'll solve some PDEs over it.

how do you make it so sexy in R

I had a R economic class and made realy ugly basic graphs

Octave is really good. gonna sound like a shill here, but i used this on my senior thesis in math so I'm gonna say how much i like it anyway. it's a free open source scientific programming environment with 99% Matlab compatible scripts. it's basically the same as personal use Matlab with no add ons. very good, i was able to write a bunch of curve fitting software in it. I'll post some good plots in it when i finish my exams today.

Use dplyr to split your data, use ggplot2 for the graph, create a bar chart, put it into polar coordinates to make it a pie chart, make it a stacked pie chart by making your alpha level and your colours represent categories, use a built in automatic colour scheme, use Cairo to do the anti-aliasing and make it look smooth.

The problem is the fact that it doesn't have the MATLAB add ons. You're better off using an open-source language with more support in that case, like Python or R.

Swastika.

oh my

...

pentagram

That's actually pretty based

alpha = (number between 0 and 1)

ggplot2 dude never use the base plots unless you r a heathen

here is the code for that graph specifically, its on the widely known mpg data set

library(ggplot2)
theme_set(theme_classic())

# Plot
g

>Gray backgrounds
This needs to die.

too lazy to add the +theme_bw() at the end, doesnt make that much of a difference, pic related

(comparing with )

The difference is better contrast, better performance in print. Gray backgrounds are the worst kind of form over function. The only reason why people think it looks good is because it looks different and new.

The plot looks shit anyway, but that's your fault.

I noticed some ppl on this board want to learn python to plot things so here is the code for that pic:

pastebin.com/7L1vZHQu

>the plot looks like shit anyway

pmsl, post a plot youve made? and youre wrong, gray is just the default background for ggplot2, as stated before in the thread this is just for meandering not finished products ready to be printed in 100000 copies of a texbook

I've been considering learning mathematica. How is it in comparison to matlab?

Any love for Gnumeric?

Pretty cool, how'd you gather the data? Did you manually take note of it all everyday, or did you have some sort of program take the measurements daily?

not him but I autistically asked him the same question the other day when he made the thread and he downloads the data from his electricity provider and just gets weather data online. P cool, but have never heard of gnumeric

My utility company lets me download my usage as a csv. They have smart meters that are internet connected.

Yeah it's primarily a Linux program. They stopped releasing a Windows version several years ago. I'm glad someone remembered my thread! Here's the graph I use to keep track of my fuel economy.

post more, and yea that was a memorable thread

how do you keep track of city vs total miles? v v interesting

Total miles is the trip odometer for each fill up. I keep track of all highway miles I drive during each tank of gas, then subtract that from total miles to get city miles, then turn it into a percentage of total miles. So say 400 miles on a tank, 100 on the highway, means 75% (300) will be city.

...

You can all fuck off with your fancy graphs

Excel master race here

Octave is good shit. Octave-forge is developing a good library of functions if you miss the MathWorks stuff.

Looks like it came from a reference book. Nice1

This is dope, cellular automaton? Mind explaining some more about it?

hand drawn autist graphs ftw

you can make base plots look exactly like ggplot with about the same amount of effort, and you don't have to use the weird grammar of ggplot

i mean i still use ggplot, but it's not inherently better than base graphics. it's just that the defaults for base graphics look pretty terrible

You reek of genuine bullshit

Stay away from Octave, it's shit. If you want to learn something useful go for Python or R.

>ITT: people with skills, but without real data

Sophistry.

No, GNU Octave is good shit.

mine was based on real data

I hereby exclude you from my predication.

I was looking for some matlab code I wrote to simulate a Belousov–Zhabotinsky reaction but I seem to have lost the code.. So here's a bifurcation diagram instead

Go on wasting your time with it then.

Go on being a poor NEET

HOHOHOHOHOHO

SILLY RABBIT

Try and make a bar graph with partial transparency.

stackoverflow.com/questions/32905859/overlap-bar-charts-with-transparency

If you have to create your own colours instead of sticking in additional arguments for alpha, that's not too great; makes it harder to scale alpha with a variable, for instance.

what are you doing this in. I've been wanting to use python to do some wave mechanics but have been too lazy. also seems like it would be a pain to set up the boundary conditions without spending the time to make a functional gui to go with it.

Thank you.

not really Veeky Forums related

made in matlab tho

dis a big graph mane

for you