Creating Arrays

There are three everyday ways to build a NumPy array. First, always import the library — the standard alias is np.

example.py

import numpy as np

1From a list — np.array()

np.array() is the fundamental function to create arrays. It takes a list (or tuple, or other array-like object) as input.

playground.py

import numpy as np

my_list = [1, 2, 3, 4]
my_array = np.array(my_list)

print(my_array)

import numpy as np

my_list = [1, 2, 3, 4]
my_array = np.array(my_list)

print(my_array)

2Evenly spaced by step — np.arange()

Similar to Python's range(), but it returns a NumPy array and — unlike range() — it happily accepts non-integer start/stop/step values.

Syntax: np.arange(start, stop, step) — stop is exclusive.

Try it — np.arange()

start0stop10step2

np.arange(0, 10, 2)

length = 5 · stop is exclusive

playground.py

import numpy as np

print(np.arange(10))          # 0 to 9
print(np.arange(2, 11, 2))    # stop is exclusive
print(np.arange(0, 5, 0.5))   # non-integer steps allowed

import numpy as np

print(np.arange(10))          # 0 to 9
print(np.arange(2, 11, 2))    # stop is exclusive
print(np.arange(0, 5, 0.5))   # non-integer steps allowed

range() vs np.arange()

Python's range(0.5, 5.5, 0.5) raises a TypeError — it only accepts integers. NumPy's np.arange(0.5, 5.5, 0.5) works perfectly and returns floats.

3Evenly spaced by count — np.linspace()

Creates evenly spaced numbers over an interval, but here you say how many values you want, rather than the step size.

Syntax: np.linspace(start, stop, num=50, endpoint=True)

Try it — np.linspace()

start0stop10num5

endpoint=True(include the stop value?)

np.linspace(0, 10, 5, endpoint=True)

2.5

7.5

You choose how many points (5); NumPy computes the spacing for you.

playground.py

import numpy as np

# 5 evenly spaced numbers from 0 to 10 (inclusive)
print(np.linspace(0, 10, 5))

# 6 evenly spaced numbers from 0 to 1 (stop excluded)
print(np.linspace(0, 1, 6, endpoint=False))

import numpy as np

# 5 evenly spaced numbers from 0 to 10 (inclusive)
print(np.linspace(0, 10, 5))

# 6 evenly spaced numbers from 0 to 1 (stop excluded)
print(np.linspace(0, 1, 6, endpoint=False))

Which one should I use?

Reach for arange when you know the step between values. Reach for linspace when you know how many values you want — perfect for plotting a smooth range of points.

4Recap

	np.array()	np.arange()	np.linspace()
Input	a list / tuple	start, stop, step	start, stop, count
You control	the values	the step size	the number of points
Stop value	—	excluded	included (by default)
Best for	known values	fixed-step sequences	fixed-count sequences

Quick Check

What does np.arange(2, 11, 2) produce?

Quick Check

You want exactly 5 points between 0 and 1. Which do you use?

← Previous

⚡ Why NumPy?

Missing Data (NaN) 🕳️