The most efficient and easiest way to convert a NumPy array to a Pandas Series is by using pd.Series() constructor. It accepts an array as an argument and returns a Series object. Ensure that the array is single-dimensional and not multi-dimensional.

What is the need for conversion?

1. The primary difference between the NumPy array and the Pandas series is that NumPy, by default, indexes each value by integer indices. In contrast, in pandas, you can explicitly define labels, which can be of type integer, string, or any other immutable type. Label indexing is more helpful for accessing data from a collection compared to integer indexing.
2. Pandas objects can easily align the data based on their indices, whereas numpy objects lack this feature.
3. Series can easily manage NaN values, where Numpy objects have limited support for them.

The process of conversion

When we convert an array to a series, pd.Series() constructor creates a new Series object in memory. It is a different object from an array object.

Then, the data is copied from the array to that object, and now the Series is an independent object with its own data.

While creating a new Series object, if you don’t specify the labels, it will make its own RangeIndex to assign an index to each element.

import numpy as np
import pandas as pd

# Creating an nd array object
numpy_array = np.array([11, 21, 31, 41, 51])

print(numpy_array)
print(type(numpy_array))

# Converting the NumPy array to a Pandas Series
pandas_series = pd.Series(numpy_array)

# Printing the Pandas Series
print(pandas_series)

# Printing the series object type
print(type(pandas_series))

Output

As shown in the above code, the numpy array’s type is “<class ‘numpy.ndarray’>”, and after the conversion, we have a Series whose type is “<class ‘pandas.core.series.Series’>”.

Converting 2D Numpy Array to Series

You cannot convert a 2D numpy array to a Series because the Series itself is not a multi-dimensional object. A Series is a single-dimensional labeled array.

If you want to get a 2D Pandas object with rows and columns, I recommend converting it into a Pandas DataFrame.

If you are stubborn like me and still want to convert it into a Series, then you have to flatten the 2D array into a 1D array and then use pd.Series() method to get Series.

import numpy as np
import pandas as pd

arr_2d = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

# Flatten the array
arr_1d = arr_2d.flatten()

# Convert to Series
series = pd.Series(arr_1d)

print(series)
print(type(series))

Output

That’s all!