Python convert dataframe to numpy array

In this post, you will learn how to convert a dataframe to numpy array in Python programming.

A DataFrame is a tabular-like data structure containing an ordered collection of columns. The each column can be of different data types, like numeric, boolean, strings, etc. It has both a row and a column index. It is basically used for analytical purposes.

NumPy is the fundamental package for scientific computing with Python. It is a highly optimized library for numerical operations. The support of NumPy makes the task more easier. Several libraries like OpenCV, SciPy, Matplotlib when combine with NumPy increase the number of weapons in your arsenal.

We can easily convert Pandas DataFrame to numpy array by using the function DataFrame.to_numpy(). This function returns the numpy ndarray when applied on the DataFrame.

DataFrame.to_numpy(dtype=None, copy=False)

Here, the first argument is the dtype to pass to numpy.asarray(). The second argument returns the boolean value that has the default value False. This function returns the numpy.ndarray as an output.

Convert the DataFrame to numpy array

Pandas is superb for taking care of your datasets, but you may discover it lacks in the statistical analysis power you need. This where it proves to be useful to convert your DataFrame to a NumPy Array.

import pandas as pd

# create a dataframe
df = pd.DataFrame(
	[[5, 17, 10],
	[15, 8, 4],
	[7, 23, 18],
	[23, 8, 11]],
	columns=['x', 'y', 'z'])

print('DataFrame\n-------------\n', df)

# convert dataframe to numpy array
num_arr = df.to_numpy()

print('\nNumpy Array\n-------------\n', num_arr)

DataFrame
-------------
     x   y   z
0   5  17  10
1  15   8   4
2   7  23  18
3  23   8  11

Numpy Array
-------------
 [[ 5 17 10]
 [15  8  4]
 [ 7 23 18]
 [23  8 11]]

Convert the DataFrame to numpy array with different datatype

Suppose we have a DataFrame with sections of different datatypes, then the returned NumPy Array comprises of elements of a single datatype. The lowest datatype of the DataFrame is considered for the datatype of the NumPy Array. The following example demonstrates this:

import pandas as pd

# create a dataframe
df = pd.DataFrame(
	[[5, 17, 10.2],
	[15, 8, 4.3],
	[7, 23, 18.5],
	[23, 8, 11.1]],
	columns=['x', 'y', 'z'])

print('DataFrame\n-------------\n', df)
print('\nDataFrame datatypes :\n', df.dtypes)

# convert dataframe to numpy array
num_arr = df.to_numpy()

print('\nNumpy Array\n-------------\n', num_arr)
print('\nNumpy Array Datatype :', num_arr.dtype)

DataFrame
-------------
     x   y     z
0   5  17  10.2
1  15   8   4.3
2   7  23  18.5
3  23   8  11.1

DataFrame datatypes :
 x      int64
y      int64
z    float64
dtype: object

Numpy Array
-------------
 [[ 5.  17.  10.2]
 [15.   8.   4.3]
 [ 7.  23.  18.5]
 [23.   8.  11.1]]

Numpy Array Datatype : float64

Related Articles