A NumPy array is an efficient, multidimensional container for storing and manipulating large datasets of homogeneous (same) data types. Although they are similar to Python lists, they are more memory-efficient and allow for high-performance mathematical and numerical operations.

Now, it’s time to create your first NumPy arrays. The most straightforward way to do this is by using the array() function, passing either a list or a tuple as its argument.

Note

You should create NumPy arrays only from lists in all the tasks throughout our course.

Let’s take a look at an example:

As you can see, these two NumPy arrays are equal.

Specifying Data Type

The data type of the array elements is defined implicitly; however, you can specify it explicitly with the dtype parameter:

The first integer array uses the default int64 data type, which is an 8-byte integer. The second array uses int8, a 1-byte integer.

Other common NumPy data types include numpy.float16, numpy.float32, and numpy.float64, which store 2-byte, 4-byte, and 8-byte floating point numbers, respectively.

Using int8 for small integers is more memory-efficient. However, be careful when choosing the data type (dtype) to avoid overflow if the numbers are too large for the specified type.

Here is a list of possible data types in NumPy: