When converting a string (a-> b) in Python, which is faster, if statement or dictionary type?

What I did with the background

I wanted to use GCP's Cloud Functions to convert the input to a pair of words each time, such as "fruit" for "apple" and "vegetable" for "strawberry". I was wondering which is faster, ** converting with if statement ** or ** preparing a dictionary type correspondence table and converting **, so I tried it. Note) In my case, it is often a cold start, so I calculate the required time including dictionary type generation.

Conclusion

** If dictionary reading occurs every time (such as AWS Lambda or GCP Cloud Functions cold start) ** The if statement is faster.

If the key is 5 characters

If the key is 10 characters (Because I used Colab, the processing speed of each time may not be stable)

** It was an overwhelming victory of the if statement. ** **

If it is not a cold start, I feel that the dictionary type is faster, so I will try it if I feel like it (If document is annoying ...)

environment

Python 3 Google Colaboratory

Verification conditions

-Execute the code to compare ** in the same cell 100 times by changing the number of keys, and compare by the average time required. ・ There are 6 patterns of 5, 10, 25, 50, 100, and 200 keys.

Verification code example

Preparation

import time
import random, string
from statistics import mean

#Random string creation (borrowed from the reference "Generate random strings using Python")
def randomname(n):
   return ''.join(random.choices(string.ascii_letters + string.digits, k=n))

range_ = 5

smp_dict = {}
key_list = []
for i in range(range_):
    key = randomname(10)
    smp_dict[key] = randomname(20)
    key_list.append(key)

#if statement creation
for i in range(range_):
    if i == 0:
        print("      if key == '{}':".format(key_list[i]))
    else :
        print("      elif key == '{}':".format(key_list[i]))
    print("          content = '{}'".format(smp_dict[key_list[i]]))

Create verification code by copying the if statement output above

rnd_key_list = smp_dict.keys()

time_list = []
for i in range(100):
    start = time.time()
    for key in rnd_key_list:
        if key == 's9MT4kkqyg':
            content = 'JpNeK8tDV1FyQfXdzTS0'
        elif key == 'PKFEikSHFM':
            content = 'RxhZqdfsCqLTuxBzIrdV'
        elif key == '04Ex4druSy':
            content = 'GmT0Jx78xPuMrOyVPuik'
        elif key == 'JUJ85l8ayb':
            content = 'xVtzS0HnHHEOHML87z85'
        elif key == 'HnqyHAIcLm':
            content = 'NRNoem9JpGUcFxvccaxD'
    time_list.append(elapsed_time)
if_mean = mean(time_list)
print ("if Mean time:{}".format(if_mean) + "[sec]")


time_list = []
for i in range(100):
    start = time.time()
    Smp_dict = {'s9MT4kkqyg': 'JpNeK8tDV1FyQfXdzTS0', 'PKFEikSHFM': 'RxhZqdfsCqLTuxBzIrdV', '04Ex4druSy': 'GmT0Jx78xPuMrOyVPuik', 'JUJ85l8ayb': 'xVtzS0HnHHEOHML87z85', 'HnqyHAIcLm': 'NRNoem9JpGUcFxvccaxD'}
    for key in rnd_key_list:
        content = Smp_dict[key]
    elapsed_time = time.time() - start
    time_list.append(elapsed_time)
dict_mean = mean(time_list)
print ("dict Mean time:{}".format(dict_mean) + "[sec]")

Explanation of verification code

smp_dict = {}
key_list = []
for i in range(100):
    key = randomname(10)
    smp_dict[key] = randomname(20)
    key_list.append(key)

-> Create a dictionary type & key list with 10 random characters as keys and 20 random characters as values

rnd_key_list = smp_dict.keys()

-> ~~ The order is not preserved when creating a dictionary type ~~, so create a new key list to make the conditions the same. I received a comment from @shiracamus that I will keep it recently!

start = time.time()
Smp_dict = {'BGM85yImoJ': 'i5GozNMNXGdfrlVgmVA0', ...}

-> Read dictionary type after starting measurement

reference

[Python] Measure and display the time required for processing Generate random strings using Python