[Tutor] How to find optimisations for code

[Peter Otten]

Pat Martin wrote:

> So should we always use sets or dictionaries if possible? Are these more
> efficient because of not keeping track of their position?

Not always. If you want to know how often one entry/char occurs in a linear 
storage like a list, a string, or a file, then you can loop over the data:

num_c = some_string.count(c)  # the loop is hidden in the count() method


pairs = (line.partition("=")[::2] for line in file)
value = next(value for key, value in pairs if key == wanted_key)

However, if you expect frequent similar questions go with a lookup table:

freq = Counter(some_string)
for c in strings.ascii_lowercase:
    print(c, "occurs", freq[c], "times")


lookup = dict(pairs)
for key in wanted_keys:
    print(key, "is", lookup.get(key, "unknown"))

Before you start measure if the effort may be worthwhile, i. e. whether the 
linear approach causes a relevant delay. 

When you're done doublecheck if you got the desired improvement, i. e. 
whether you optimised the right place.