NumPy Data Types

Every NumPy array has exactly one data type, called its dtype. The dtype decides how much memory each value uses and which operations are allowed.

Check the data type

Example

import numpy as np

ints = np.array([1, 2, 3])
floats = np.array([1.5, 2.0, 3.25])
words = np.array(["apple", "kiwi"])

print(ints.dtype)
print(floats.dtype)
print(words.dtype)

Output

int64
float64
<U5

<U5 means a Unicode string of up to 5 characters. The number in names like int64 is the size of each value in bits.

Common data types

Choose a data type

Pass the dtype argument to create an array of a specific type:

Example

import numpy as np

arr = np.array([1, 2, 3], dtype="float64")
print(arr)
print(arr.dtype)

Output

[1. 2. 3.]
float64

Convert with astype()

astype() returns a new array with a different type. Converting floats to integers drops the decimal part — it does not round.

Example

import numpy as np

prices = np.array([9.99, 15.5, 3.2])
print(prices.astype(int))

flags = np.array([0, 1, 2, 0])
print(flags.astype(bool))

Output

[ 9 15  3]
[False  True  True False]

Mixed values are upcast

If you mix types, NumPy picks one type that can hold all of them. Integers and floats together become floats:

Example

import numpy as np

mixed = np.array([1, 2, 3.5])
print(mixed)
print(mixed.dtype)

Output

[1.  2.  3.5]
float64

If a value cannot be converted, NumPy raises a ValueError — for example np.array(['a', '2'], dtype=int).