Example code for in-memory proximity search with Python (2).

gistfile1.py

import flask
from flask.ext.cache import Cache
from siren import util, crime_tracker
from siren.decorators import jsonp


app = flask.Flask(__name__)
app.config.from_envvar('SIREN_SETTINGS', silent=True)
cache = Cache(app)
crime_db = crime_tracker.PortlandCrimeTracker()


def get_point(latitude, longitude):
    """
    Return ``latitude`` and ``longitude`` coerced to floats.
    Fail the request with a 400 status if one or both are not coercible.
    """
    try:
        latitude = float(latitude)
        longitude = float(longitude)
    except ValueError:
        flask.abort(400)

    return latitude, longitude


def get_crimes(latitude, longitude):
    """
    Return all crimes found within 1/2 mile of ``latitude`` and ``longitude``.
    """
    point = get_point(latitude, longitude)
    ignore = ['callback', '_']
    filters = {k: v for k, v in flask.request.args.items() if k not in ignore}
    return crime_db.get_crimes_nearby(point, filters)


@app.route('/crime/stats/<latitude>,<longitude>')
@cache.cached()
@jsonp
def crime_stats(latitude, longitude):
    """
    Return statistics -- right now, only sums by category -- about crimes
    within 1/2 mile of ``latitude`` and ``longitude``.
    """
    nearby_crimes, errors = get_crimes(latitude, longitude)
    stats = crime_db.get_stats_for_crimes(nearby_crimes)
    data = {'stats': stats}

    if errors:
        data['errors'] = errors

    return flask.jsonify(result=data)


@app.route('/crime/<latitude>,<longitude>')
@cache.cached()
@jsonp
def crimes(latitude, longitude):
    """
    Return raw crime data for all crimes within 1/2 mile of ``latitude`` and
    ``longitude``.
    """
    crimes, errors = get_crimes(latitude, longitude)
    # JSON requires that object keys be strings, so convert tuples to strings
    # like "37.785834,-122.406417"
    crimes = [dict(crime=c, point=','.join(str(p))) for p, c in crimes.items()]
    data = {'crimes': crimes}

    if errors:
        data['errors'] = errors

    return flask.jsonify(result=data)

Hello,
I just found your code and I am trying to do a very similar task, but I am having a lot of issues with reconstructing the kd-tree using a task queue for distributed processing, as this is in-memory data structure.
I would really much appreciate if you can have a short-look at this question I askedon stackoverflow.
http://stackoverflow.com/questions/26088868/asynchronous-task-queue-processing-of-in-memory-data-stucture-in-django?noredirect=1#comment40883596_26088868

Thanks a lot