These are provided by the numpy module - so you will need to import it! if you use the traditional numpy import
import numpy as np
Then you will need to prefix all functions with np.
Note that there also a few methods of the array class - see numpy array methods for these.
Functions are given as:
**function_name(arguments) return type: return_type
arguments is the list of arguments the function expects – the text will tell you what they do. I haven’t always documented all arguments.
return_type is the type of value the function returns – in many cases this is a string or a number. Some functions do not return any value – these are given as return type: N/A
linspace(start, stop, count) return type: array
linspace returns a numpy array of count evenly spaced numbers between (and including) the start and stop values.
For example, linspace(0,5,10) returns [0 0.625 1.25 1.875 2.5 3.125 3.75 4.375 5]. linspace is similar in some ways to the built-in range function,
but returns floating point numbers not integers. It has all sorts of uses – one is in plotting graphs.
If you have a function that that takes a value x and returns y, just use linspace to generate a range of x values,
and then use these to generate your y values. NOTE – unlike range, the ‘stop’ value for linspace IS the last value generated (range stops at the stop value – 1).
transpose(array) return type: array Transposes a multidimensional array (swaps rows and columns]. No effect on one-dimensional arrays.
zeros(size) return type: array
zeros (make sure you don’t spell it zeroes) creates an array with elements initialised to zero (as a floating point type).
To create a 1-dimensional array (a vector), just use a number for size, e.g. zeros(3) to create [0 0 0].
To create a multidimensional array, pass a tuple as an argument, specifying the sizes of each dimension of the array.
So for a two-dimensional array (a table) with sizes 10 and 5, use zeros((10,5)). Note the double-brackets –
this is because the argument is a tuple.
Mathematical functions: Numpy includes a lot of maths functions, including pretty well everything that’s in the math module, and some more besides. You need the numpy version if you are going to use them with vectorization. I’m not listing these here - just look them up, e.g. at (https://www.geeksforgeeks.org/numpy-mathematical-function/)[https://www.geeksforgeeks.org/numpy-mathematical-function/]. A lot of these (e.g. sin) have the same name as functions in the math module – if you import both, avoid the import * syntax! You have been warned…