Updated March 23, 2023
Overview of NumPy Functions
Numpy is a python package used for scientific computing. So certainly, it supports a vast variety of functions used for computation. The various functions supported by numpy are mathematical, financial, universal, windows, and logical functions. Universal functions are used for array broadcasting, typecasting, and several other standard features. While windows functions are used in signal processing. We will be learning mathematical functions in detail in this article.
Mathematical Functions in NumPy
Numpy is written purely in C language. Hence, it’s mathematical functions are closely associated with functions present is math.h library in C.
1. Arithmetic Functions
|
Function |
Description |
| add(arr1, arr2,..) | Add arrays element wise |
| reciprocal(arr) | Returns reciprocal of elements of the argument array |
| negative(arr) | Returns numerical negative of elements of an array |
| multiply(arr1,arr2,…) | Multiply arrays element wise |
| divide(arr1,arr2) | Divide arrays element wise |
| power(arr1,arr2) | Return the first array with its each of its elements raised to the power of elements in the second array (element wise) |
| subtract(arr1,arr2,…) | Subtract arrays element wise |
| true_divide(arr1,arr2) | Returns true_divide of an array element wise |
| floor_divide(arr1,arr2) | Returns floor after dividing an array element wise |
| float_power(arr1,arr2) | Return the first array with its each of its elements raised to the power of elements in the second array (elementwise) |
| fmod(arr1,arr2) | Returns floor of the remainder after division elementwise |
| mod(arr1,arr2) | Returns remainder after division elementwise |
| remainder(arr1,arr2) | Returns remainder after division elementwise |
| divmod(arr1,arr2) | Returns remainder and quotient after division elementwise |
The above-mentioned operations can be performed in the following examples:
Example #1
In the given code snippet, we try to do some basic operations on the arguments, array a and array b.
Code:
import numpy as np
a = np.array([10,20,30])
b= np.array([1,2,3])
print("addition of a and b :",np.add(a,b))
print("multiplication of a and b :",np.multiply(a,b))
print("subtraction of a and b :",np.subtract(a,b))
print("a raised to b is:",np.power(a,b))Output:
Example #2
In this code snippet, we try to perform division and related operations on the arguments, array a and array b. We can notice the difference between mod, remainder, divmod and simple division.
Code:
import numpy as np
a = np.array([10,20,30])
b= np.array([2,3,4])
print("division of a and b :",np.divide(a,b))
print("true division of a :",np.true_divide(a,b))
print("floor_division of a and b :",np.floor_divide(a,b))
print("float_power of a raised to b :",np.float_power(a,b))
print("fmod of a and b :",np.fmod(a,b))
print("mod of a and b :",np.mod(a,b))
print("quotient and remainder of a and b :",np.divmod(a,b))
print("remainders when a/b :",np.remainder(a,b))Output:
2. Trigonometric Functions
| Function | Description |
| sin(arr) | Returns trigonometric sine element wise |
| cos(arr) | Returns trigonometric cos element wise |
| tan(arr) | Returns trigonometric tan element wise |
| arcsin(arr) | Returns trigonometric inverse sine element wise |
| arccos(arr) | Returns trigonometric inverse cosine element wise |
| arctan(arr) | Returns trigonometric inverse tan element wise |
| hypot(a,b) | Returns hypotenuse of a right triangle with perpendicular and base as arguments |
| degrees(arr)
rad2deg(arr) |
Covert input angles from radians to degrees |
| radians(arr)
deg2rad(arr) |
Covert input angles from degrees to radians |
Here is an example of how to use trigonometric functions.
Example
Code:
import numpy as np
angles = np.array([0,np.pi/2, np.pi]) -----> input array angles
sin_angles = np.sin(angles)
cosine_angles = np.cos(angles)
tan_angles = np.tan(angles)
rad2degree = np.degrees(angles)
print("sin of angles:",sin_angles)
print("cosine of angles:",cosine_angles)
print("tan of angles:",tan_angles)
print("angles in radians",rad2degree)Output:
3. Logarithmic and Exponential Functions
| Function | Description |
| exp(arr) | Returns exponential of an input array element wise |
| expm1(arr) | Returns exponential exp(x)-1 of an input array element wise |
| exp2(arr) | Returns exponential 2**x of all elements in an array |
| log(arr) | Returns natural log of an input array element wise |
| log10(arr) | Returns log base 10 of an input array element wise |
| log2(arr) | Returns log base 2 of an input array element wise |
| logaddexp(arr) | Returns logarithm of the sum of exponentiations of all inputs |
| logaddexp2(arr) | Returns logarithm of the sum of exponentiations of the inputs in base 2 |
Here is an example of using logarithmic functions:
Example
Code:
import numpy as np
a = np.array([1,2,3,4,5])
a_log = np.log(a)
a_exp = np.exp(a)
print("log of input array a is:",a_log)
print("exponent of input array a is:",a_exp)Output:
4. Rounding Functions
| Function | Description |
| around(arr,decimal) | Rounds the elements of an input array upto given decimal places |
| round_(arr,decimal) | Rounds the elements of an input array upto given decimal places |
| rint(arr) | Round the elements of an input array to the nearest integer towards zero |
| fix(arr) | Round the elements of an input array to the nearest integer towards zero |
| floor(arr) | Returns floor of input array element wise |
| ceil(arr) | Returns ceiling of input array element wise |
| trunc(arr) | Return the truncated value of an input array element wise |
Example of using rounding functions with numpy arrays:
Example #1
Code:
import numpy as np
a = np.array([1.23,4.165,3.8245])
rounded_a = np.round_(a,2)
print(rounded_a)Output:
Example #2
Code:
floor_a = np.floor(a)
print(floor_a)Output:
5. Miscellaneous Functions
| Function | Description |
| sqrt(arr) | Returns the square root of an input array element wise |
| cbrt(arr) | Returns cube root of an input array element wise |
| absolute(arr) | Returns absolute value each element in an input array |
| maximum(arr1,arr2,…) | Returns element wise maximum of the input arrays |
| minimum(arr1,arr2,…) | Returns element wise minimum of the input arrays |
| interp(arr, xp, fp) | Calculates one-dimensional linear interpolation |
| convolve(arr, v) | Returns linear convolution of two one-dimensional sequences |
| clip(arr, arr_min, arr_max) | Limits the values in an input array |
Some examples using the above functions.
Example #1 – Finding the Maxima
Code:
import numpy as np
a = [1,2,3]
b = [3,1,2]
maximum_elementwise = np.maximum(a,b)
print("maxima are:",maximum_elementwise)Output:
Example #2 – Clipping an array between max_limit and min_limit
Code:
import numpy as np
a = [1,2,3]
b = [3,1,2]
limiting_a = np.clip(a,0,2)
print("limiting a between 0 and 2:",limiting_a)Output:
Conclusion
In this post, we have tried to cover the most frequently used mathematical functions in numpy. However, there are some other functions like complex functions which helps in working with real and imaginary parts of a number, floating-point routines which helps in performing decimal operations and hyperbolic functions which helps in calculating hyperbolic sine, tan, cos, etc.
Recommended Articles
This is a guide to NumPy Functions. Here we discuss the basic concept and different mathematical functions in NumPy. You may also look at the following articles to learn more –
