# Using numpy hstack() to horizontally stack arrays

In this tutorial, we will look at how to use the numpy hstack method to horizontally stack (or concat) numpy arrays with the help of some examples.

You can use the numpy hstack() function to stack numpy arrays horizontally. It concatenates the arrays in sequence horizontally (column-wise). The following is the syntax.

```import numpy as np

# tup is a tuple of arrays to be concatenated, e.g. (ar1, ar2, ..)
ar_h = np.hstack(tup)```

It takes the sequence of arrays to be concatenated as a parameter and returns a numpy array resulting from stacking the given arrays. This function is similar to the numpy vstack() function which is also used to concatenate arrays but it stacks them vertically.

Let’s look at some examples of how to use the numpy hstack() function.

Let’s stack two one-dimensional arrays together horizontally.

```import numpy as np

# create two 1d arrays
ar1 = np.array([1, 2, 3])
ar2 = np.array([4, 5, 6])
# hstack the arrays
ar_h = np.hstack((ar1, ar2))
# display the concatenated array
print(ar_h)```

Output:

`[1 2 3 4 5 6]`

Here, we created two 1D arrays of length 3 and then horizontally stacked them with the hstack() function. The resulting array is also one-dimensional since we are concatenating them horizontally.

You can also stack more than two arrays at once with the numpy hstack() function. Just pass the arrays to be stacked as a tuple. For example, let’s stack three 1D arrays horizontally at once.

```# create two 1d arrays
ar1 = np.array([1, 2, 3])
ar2 = np.array([4, 5, 6])
ar3 = np.array([7, 8, 9])
# hstack the arrays
ar_h = np.hstack((ar1, ar2, ar3))
# display the concatenated array
print(ar_h)```

Output:

`[1 2 3 4 5 6 7 8 9]`

Here we concatenated three arrays of length 3 horizontally. The resulting array is one-dimensional and has length 9. Similarly, you can stack multiple arrays, just pass them in the order you want as a sequence.

Now let’s try to stack a 1D array with a 2D array horizontally.

```# create a 1d array
ar1 = np.array([1, 2])
# create a 2d array
ar2 = np.array([[0, 0, 0],
[1, 1, 1]])

# hstack the arrays
ar_h = np.hstack((ar1, ar2))
# display the concatenated array
print(ar_h)```

Output:

```---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-6-5f5f4f1cdb7b> in <module>
6
7 # hstack the arrays
----> 8 ar_h = np.hstack((ar1, ar2))
9 # display the concatenated array
10 print(ar_h)

<__array_function__ internals> in hstack(*args, **kwargs)

~\anaconda3\envs\dsp\lib\site-packages\numpy\core\shape_base.py in hstack(tup)
341     # As a special case, dimension 0 of 1-dimensional arrays is "horizontal"
342     if arrs and arrs.ndim == 1:
--> 343         return _nx.concatenate(arrs, 0)
344     else:
345         return _nx.concatenate(arrs, 1)

<__array_function__ internals> in concatenate(*args, **kwargs)

ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 2 dimension(s)```

We get an error. This is because the dimensions of the arrays are such that you cannot stack them together horizontally.

If you want to stack the two arrays horizontally, they need to have the same number of rows.

```# create an array of shape (2, 1)
ar1 = np.array([, ])
# create a 2d array
ar2 = np.array([[0, 0, 0],
[1, 1, 1]])

# hstack the arrays
ar_h = np.hstack((ar1, ar2))
# display the concatenated array
print(ar_h)```

Output:

``` [[1 0 0 0]
[2 1 1 1]]```

You can also horizontally stack two 2D arrays together in a similar way.

```# create a 2d array
ar1 = np.array([[1, 2],
[3, 4],
[5, 6]])
# create a 2d array
ar2 = np.array([[0, 0],
[1, 1],
[2, 2]])

# hstack the arrays
ar_h = np.hstack((ar1, ar2))
# display the concatenated array
print(ar_h)```

Output:

```[[1 2 0 0]
[3 4 1 1]
[5 6 2 2]]```

Here we stacked two 2d arrays of shape (3, 2) horizontally resulting in an array of shape (3, 4).

For more on the numpy hstack() function, refer to its documentation.

With this, we come to the end of this tutorial. The code examples and results presented in this tutorial have been implemented in a Jupyter Notebook with a python (version 3.8.3) kernel having numpy version 1.18.5

• 