Implemented performance improvement to access scores API call.

[davphan]

Resolves #2988

Before, the computation for the access streets score was done as follows:

for every street s:
    for every clustered attribute c:
        create a jts Point object from the c's latitude and longitude, Point(c)
        if the Point(c) is within 10 meters of s:
            increase cluster count for s

which duplicated the Point creation step for all labels on every street. The computation was modified to follow this scheme:

for every clustered attribute c:
    create a jts Point object from the c's latitude and longitude, Point(c)
for every street s:
    for every Point(c):
        if Point(c) is within 10 meters of s:
            increase cluster count for s

to reduce the amount of creations of Point objects to just one per label.

Testing instructions

1. Go to the API page (http://localhost:9000/api) and use the drop down in the "Access Score: Streets" section to get the scores for the entire city (or use this link, http://[localhost:9000/v2/access/score/streets?filetype=geojson) and verify that the API call still works.

Things to check before submitting the PR

• I've written a descriptive PR title.
• I've added/updated comments for large or confusing blocks of code.

[misaugstad]

We talked about this during the meeting today, but as a reminder:

1. Can you check that the output is exactly the same?
2. Can you check what sort of performance improvement you get? I think that you have the full Pittsburgh database that is pretty recent, so I'm hoping that there is enough data to be able to measure a difference. Maybe run it 5 times or so with each version to get an average time for the computation!

[davphan]

Performance benchmarks:
Runtime calculated using System.nano() at the beginning and end of the computeAccessScoresForStreets method, averaging 5 runs with no caching.
Before update: 7.62 s average runtime
After update: 2.87 s average runtime

Output difference:
Checked for any differences between the geojson file output.
Exactly the same, order and data wise!

[misaugstad]

Awesome, thank you @davphan! Will test (and I'm assuming merge) tomorrow!

[misaugstad]

Nicely done!!