Problem Formulation and Solution Overview
To make it more interesting, we have the following running scenario:
Carrie heard that Creative Prints is hiring a Python coder. They require the interviewee to answer several coding questions: one is to provide several ways to find the most common element in a NumPy Array.
Preparation
Before moving forward, please ensure that the NumPy library is installed.
import numpy as np
Method 1: Use bincount() and argmax()
This example uses bincount() to tally the occurrences of each element in an array of positive integers and argmax() to return the maximum values along an axis.
max_value = np.bincount([4, 18, 2, 8, 3, 15, 14, 15, 20, 12, 6, 3, 15, 12, 13, 19, 14, 81, 23, 44]).argmax() print(max_value)
The above code calls in NumPy’s bincount() function and passes it one (1) argument, a random list of positive integers.
Then, argmax() is called to retrieve the maximum value in the list . The results save to max_value and are output to the terminal.
| 15 |
The output is correct! This number occurs 3 times in the list.
💡Note: This option does not work with negative numbers or strings.
Method 2: Use Collections.Counter
This example uses the collections library that provides container types allowing the coder to easily store and access data values. The Counter() function keeps a running tally of each element’s count.
from collections import Counter values = np.array([4, 18, 2, 8, 3, 5, 14, 5, -81, 12, 6, 3, -81, 12, 13, -81, 14, -81, 23, 44]) max_value = Counter(values) print(max_value.most_common(1))
The above code imports Python’s built-in collections library and the Counter()
Then, a NumPy Array is declared containing a list of positive and negative random numbers. This saves to values.
Next, the collections.Counter() is called and passed the variable values.max_value. If output to the terminal, at this point, max_value would contain the following.
Counter({-81: 4, 3: 2, 5: 2, 14: 2, 12: 2, 4: 1, 18: 1, 2: 1, 8: 1, 6: 1, 13: 1, 23: 1, 44: 1}) |
Finally, most_common (a counter tool for quick tallies) is appended to max_value
A List of Tuples containing the number with the highest tally and the associated count returns and is output to the terminal.
In this case, -81 occurred the most times with 4 occurrences.
[(-81, 4)] |
To break the Tuple out of the List, append [0] to the end.
print(max_value.most_common(1)[0])
This will result in the following output.
(-81, 4) |
Method 3: Use unique()
This example uses NumPy’s unique() function that sorts and returns the unique array elements with their respective counts in a List format.
values = np.array([4, 18, 2, 8, 3, 5, 14, 5, -81, 12, 6, 3, -81, 12, 13, -81, 14, -81, 23, 44]) vals, counts = np.unique(values, return_counts=True) print(vals) print(counts)
The above code declares a NumPy Array containing a list of positive and negative random numbers. This saves to values.
Next, unique() is called and passed two (2) arguments: the values list created above and return_counts. If set to True, this function will return the unique values and their respective occurrences. The results save to vals and counts.
The contents of vals and counts are output to the terminal using separate print() statements.
[-81 2 3 4 5 6 8 12 13 14 18 23 44] |
To retrieve the most comment element and its respective count, use slicing.
print(vals[0], counts[0])
The output from the above is as follows.
-81 4 |
Method 4: Use unique() and zip()
This example uses NumPy’s unique() and the zip() function, which takes an arbitrary number of iterables and returns a Dictionary based on number of occurrences as the key and total count as the value.
values = np.array([4, 18, 2, 8, 3, 5, 14, 5, -81, 12, 6, 3, -81, 12, 13, -81, 14, -81, 23, 44]) result = dict(zip(*np.unique(values, return_counts=True))) print(result)
The above code declares a NumPy Array containing a list of positive and negative random numbers. This saves to values.
Then, zip() is called and passed one (1) argument, unique(), which passes two (2) arguments: the values list created above and return_counts. If return_counts is set to True, this will return the unique values and their respective occurrences.
Next, dict() is called which merges the data above into a Dictionary format. This saves to result and is output to the terminal.
{-81: 4, 2: 1, 3: 2, 4: 1, 5: 2, 6: 1, 8: 1, 12: 2, 13: 1, 14: 2, 18: 1, 23: 1, 44: 1} |
To retrieve the most common element, run the following code.
print(list(result.items())[0:1])
The output from the above is as follows.
[(-81, 4)] |
Method 5: Use mode()
This example uses mode() from the statistics library. This function returns the single most common element found in the passed argument.
from statistics import mode
def most_common(List):
return(mode(List))
values = np.array([4, 18, 2, 8, 3, 5, 14, 5, -81, 12, 6, 3, -81, 12, 13, -81, 14, -81, 23, 44])
print(most_common(values))
The above code calls in mode() from the statistics library.
Then, a definition is defined called most_common and accepts one (1) argument, a list. The mode() function is applied to the list and is returned.
Next, a NumPy Array is declared containing random numbers and saves to values.
The above most_common function is called, passing the NumPy Array values and output to the terminal.
| -81 |
Summary
This article has provided five (5) ways to find the most common element in a NumPy Array. These examples should give you enough information to select the best fitting for your coding requirements.
Good Luck & Happy Coding!
Programming Humor – Python
At university, I found my love of writing and coding. Both of which I was able to use in my career.
During the past 15 years, I have held a number of positions such as:
In-house Corporate Technical Writer for various software programs such as Navision and Microsoft CRM
Corporate Trainer (staff of 30+)
Programming Instructor
Implementation Specialist for Navision and Microsoft CRM
Senior PHP Coder