Skip to content

cas.md

Latest commit

 

History

History
201 lines (151 loc) · 5.39 KB

cas.md

File metadata and controls

201 lines (151 loc) · 5.39 KB
description Introduction to `cas-rs`, the upcoming new computer algebra system

Introduction

cas-rs is the new computer algebra system that will eventually power CalcBot, including the c-calculate command, and practically every other command that involves math. It's designed to be significantly faster, more accurate and more feature-rich than the current system. Even if you don't need or use the new features, you'll still benefit from very fast calculations.

What's the -rs for?

-rs is often added to the names of projects written in the Rust programming language. You can browse all the code for cas-rs here on GitHub.

All you really want to know is that Rust makes cas-rs really fast.

Usage

You can try cas-rs right now in CalcBot by enabling preview mode, then using the c-calculate command as normal. It will function just like the previous c-calculate, but your input is fed to cas-rs instead of the current system, except the calculation will be noticeably faster.

> c-calculate 0.45/240
0.001875

> c-calculate sin(2pi)
0

> c-calculate 0.279*3.11
0.86769

Improvements over the current system

cas-rs introduces many new features and bug fixes that make calculations much easier and more powerful.

Better errors

When an error occurs in cas-rs, you'll now get a neat diagram explaining exactly what went wrong, and in most cases, how to fix it. Here's an example:

> c-calculate log()
Error: missing required argument #1 for the `log` function
   ╭─[input:1:1]
   │
 1 │ log( )
   │ ──┬─ ─
   │   ╰───── this function call
   │
   │ Help: the `log` function takes 2 argument(s); there are 0 argument(s) provided here
   │
   │ Note: function signature: `log(n: Complex, base: Complex (optional))

Compare this to the current system's error message:

> c-calculate log()
log()
     
The first argument of the `log` function must be of type `number` or `complex`.
Evaluated to: undefined

Lists

Lists are a new type of data in cas-rs, which let you group multiple values together in a sequence. Lists are created using square brackets [], with values inside, separated by commas:

> c-calculate list = [1, 2, 3, 4, 5]
[1, 2, 3, 4, 5]

Perhaps the most useful feature of lists is that you can perform operations on all elements of a list at once. For example, to convert multiple temperatures from Celsius to Fahrenheit in one go:

> c-calculate c = [0, 20, 37, 100]
[0, 20, 37, 100]

> c-calculate f = c * 9/5 + 32
[32, 68, 98.6, 212]

This extends to nested lists as well, allowing you to simulate operations on matrices, and more:

> c-calculate matrix = [[1, 2], [3, 4]]
[[1, 2], [3, 4]]

> c-calculate 2matrix
[[2, 4], [6, 8]]

Conditionals

cas-rs introduces conditional expressions, which let you choose between two or more values based on various conditions. This was previously possible with the if function, but the new syntax is significantly more readable and programming-like. For example, to calculate whether a year is a leap year or not:

> c-calculate year = 2024
2024

> c-calculate isleapyear = {
    if year % 400 == 0 {
        true
    } else if year % 100 == 0 {
        false
    } else if year % 4 == 0 {
        true
    } else {
        false
    }
}
true

The old way of doing this using the if function was much more difficult to read. See for yourself:

> c-calculate year = 2024
2024

> c-calculate isleapyear = if(year % 4 == 0, if(year % 100 == 0, if(year % 400 == 0, true, false), true), false)
1

Functions

Functions let you encapsulate a piece of logic or calculation, and reuse it multiple times. In cas-rs, functions are now truly real functions! They can span multiple lines and contain complex logic, including anything introduced here.

Loops

cas-rs introduces loops, which let you repeat a block of code / calculation multiple times.

A for loop runs a calculation a specific number of times. Here's an example of a function that performs Newton's method to solve a given equation. It uses prime notation (f'(guess)), which has actually existed in CalcBot for a while:

> c-calculate newton(f, guess) = {
    for i in 0..10 {
        guess -= f(guess) / f'(guess)
    }
    guess
}; newton(f(x) = x^2 + 5x + 6, 0)
-2

for in in 0..10 means "repeat the following code 10 times, with i taking values from 0 to 9". The loop variable i can be used within the

Summation and product notation

There is a new syntax for computing the sum or product of a sequence, using the sum and product keywords. This was previously possible using the sum and product functions, however the new syntax is more concise and easier to read.

> c-calculate sum n in 1..=100 of n
5050

Assignment actually works

Function and variable assignment was actually very broken in the current system. The system to detect assignments was very fragile and could warp the input in unexpected ways. Try this in c-calculate:

> c-calculate 3 / (8) == 3 / 8
ERROR: `3 /` cannot be used as a function name.

> c-calculate (35-b)+b+(32-b)==78
ERROR: `(35-b)+b+` cannot be used as a function name.

cas-rs fixes this, so you can now assign variables and functions and use them within expressions.