Performance deep dive

In this approach, we'll find out how to most efficiently determine if a string is an Isogram in Python.

The approaches page lists four idiomatic approaches to this exercise:

1. Using a list comprehension and set
2. Using replace and set
3. Using a re.sub and set
4. Using a re.findall and set

For our performance investigation, we'll also include a fifth approach that uses a bit field to keep track of used letters.

Benchmarks

To benchmark the approaches, we wrote a small benchmark application using the timeit library.

scrubbed comprehension: 3.118929599993862e-06
scrubbed replace:       9.586393000063253e-07
scrubbed regex:         1.8171838999987813e-06
findall regex:          4.059006099996623e-06
bitfield:               5.4183307999919636e-06

The four fastest approaches use set.

• Calling a series of replace methods to scrub the input was the fastest approach at about 959 nanoseconds.
• Next fastest was using re.sub to scrub the input, at about 1817 nanoseconds.
• Third fastest was using a list comprehension to scrub the input, at about 3119 nanoseconds.
• Using re.findall to scrub the input was fourth fastest, at about 4059 nanoseconds.
• Although the bit field approach may be faster in other languages, it is significantly slower in Python. It was slower than all of the set approaches, at about 5418 nanoseconds.