Showing a way of returning elementwise operations to some functions

[dvd101x]

Hi everyone,

I have been working on a way to extend some mathjs functions to add elementwise operations similar to matlab or numpy (personally I'm more familiar with that)

trigonometric functions, exp, log, gamma, square, sqrt, cube, and cbrt

To test it:

• Open this site https://dvd101x.github.io/web-thermodynamics/
• Paste these commands on the left pane.

x = 1:3

exp(x)
log(x)
log(x,2)
gamma(x)
square(x)
cube(x)
cbrt(x)

# Trigonometrics
acos(x)
acosh(x)
acot(x)
acoth(x)
acsc(x)
acsch(x)
asec(x)
asech(x) 
asin(x)
asinh(x) 	
atan(x)
atan2(x, x)
atanh(x)
cos(x)
cosh(x)
cot(x)
coth(x)
csc(x)
csch(x)
sec(x)
sech(x)
sin(x)
sinh(x)
tan(x)
tanh(x)

It's the first time I use some of the extension capabilities of mathjs and I'm fascinated by that. Maybe it's a bit basic for some of the more advanced users here and I see those capabilities are well documented, nontheless here it is.

Even found a way to make work log([1, 2], 2) logarithm base 2 elementwise for [1, 2]

https://github.com/dvd101x/web-thermodynamics

PD: Next I will try to extend index

[josdejong]

Thanks for sharing David!

You can indeed extend a mathjs function with support for Array and Matrix, for example to extend the function sin:

// extend function sin with support for matrices and arrays
math.import({
  sin: math.typed({
    'Array | Matrix': A => math.map(A, math.sin)
  })
})

For reference: there is a discussion whether to support element-wise operations for a set of functions in #2440

[dvd101x]

Thank you Jos!

I will reade more about math.typed( ) as I currently don't undertand it but seems to be very useful for this and other stuff I'm testing.

[josdejong]

👍 you can read more about it on the following page: https://mathjs.org/docs/core/extension.html#typed-functions

[dvd101x]

Thank you!

I’m starting to understand it and I’m making a few tests. I will try to update my code accordingly.

[dvd101x]

I used it extensively and it worked like a charm.

Thank you so much!