def totalTweets(bin, rdd):
lPointObj = []
for cPoint in bin.poly:
lPointObj.append(pointClass.Point(cPoint[1], cPoint[0]))
clusterPoly = pointClass.Polygon(lPointObj)
bin.postsInHull = rdd.filter(lambda x: clusterPoly.bPointInPoly(
pointClass.Point(x["longitude"], x["latitude"]))).count()
def makeTotalsArray(record, bc_lClustPoly, bUseTime):
lRet = []
for poly in bc_lClustPoly.value:
if not bUseTime and poly.bPointInPoly(
pointClass.Point(record.lon, record.lat)):
lRet.append(1)
elif bUseTime and poly.bEventInTime(
pointClass.Point(record.lon, record.lat), record.dt.date()):
lRet.append(1)
else:
lRet.append(0)
return lRet
def testListFromData(data):
#return value is list of lists;
# 0 - polygons labeldd 'train'
# 1 - polygons labeled 'pos'
# 2 - polygons labeled 'neg'
lTrain, lPos, lNeg = [], [], []
for site in data['sites']:
pList = []
dList = []
for i2 in range(len(site['lats'])):
pList.append(pointClass.Point(site['lons'][i2], site['lats'][i2]))
for datePairs in site['dates']:
mindate = datetime.strptime(datePairs['min'].split('T')[0],
'%Y-%m-%d').date()
maxdate = datetime.strptime(datePairs['max'].split('T')[0],
'%Y-%m-%d').date()
dList.append((mindate, maxdate))
name = site["name"]
if name == "train":
lTrain.append(pointClass.spaceTimePlane(pList, dList, name))
elif name == "pos":
lPos.append(pointClass.spaceTimePlane(pList, dList, name))
elif name == "neg":
lNeg.append(pointClass.spaceTimePlane(pList, dList, name))
else:
print "Polygon name does not follow correct scheme, (train, pos, neg)"
return (lTrain, lPos, lNeg)
def outEOI(lat, lon, dt, bc_lTargetPolygons):
recordPoint = pointClass.Point(lon, lat)
rDate = dt.date()
for site in bc_lTargetPolygons.value:
if site.bEventOutOfTime(recordPoint, rDate):
return True
return False
def createHull(cluster, bUseTime):
loLa = set([])
for point in cluster.records:
loLa.add(str(point.lon) + "," + str(point.lat))
loLa = map(lambda x: [float(x.split(",")[0]),
float(x.split(",")[1])], loLa)
if len(loLa) > 2:
loLa = np.array(loLa)
hull = ConvexHull(loLa)
lClustPoints = []
lPointObj = []
for verts in hull.vertices:
lClustPoints.append([loLa[verts, 1], loLa[verts, 0]])
lPointObj.append(pointClass.Point(loLa[verts, 0], loLa[verts, 1]))
if bUseTime:
cluster.objPoly = pointClass.spaceTimePlane(
lPointObj, [(cluster.dt.date(), cluster.dt.date())], "Hull")
else:
cluster.objPoly = pointClass.Polygon(lPointObj)
cluster.poly = lClustPoints
else:
lPointObj = []
lClustPoints = []
for lolaOff in [
(-0.001, -0.001),
(-0.001, 0.001),
(0.001, 0.001),
(0.001, -0.001),
]:
offset = [loLa[0][0] + lolaOff[0], loLa[0][1] + lolaOff[1]]
lPointObj.append(pointClass.Point(offset[0], offset[1]))
lClustPoints.append([offset[1], offset[0]])
if bUseTime:
cluster.objPoly = pointClass.spaceTimePlane(
lPointObj, [(cluster.dt.date(), cluster.dt.date())], "Hull")
else:
cluster.objPoly = pointClass.Polygon(lPointObj)
cluster.poly = lClustPoints
def readInShapeDocument(data):
lST = []
siteList = data['sites']
for site in siteList:
pList = []
dList = []
for i2 in range(len(site['lats'])):
pList.append(pointClass.Point(site['lons'][i2], site['lats'][i2]))
for datePairs in site['dates']:
mindate = datetime.strptime(datePairs['min'].split('T')[0],
'%Y-%m-%d').date()
maxdate = datetime.strptime(datePairs['max'].split('T')[0],
'%Y-%m-%d').date()
dList.append((mindate, maxdate))
lST.append(pointClass.spaceTimePlane(pList, dList, site['name']))
return lST
def readInShapeFile(fileName):
pointReader = open(fileName, 'rb')
polyList = []
pointList = []
polyIndex = -1
for row in pointReader:
fInd, lat, lon = row.split(",")
if int(fInd) != polyIndex:
if polyIndex != -1:
newPoly = pointClass.Polygon(pointList)
polyList.append(newPoly)
pointList = []
polyIndex = int(fInd)
newPoint = pointClass.Point(lon, lat)
pointList.append(newPoint)
newPoly = pointClass.Polygon(pointList)
polyList.append(newPoly)
return polyList
polyFiles = os.listdir(strDir)
retDict = {}
if strDir[-1] != '/':
strDir = strDir + '/'
for file in polyFiles:
retDict[file[0:8]] = readInShapeFile(strDir + file)
return retDict
if __name__ == "__main__":
polyList = readInShapeFile("../polygon/shapeFile2.txt")
print "The capture boxes are defined by: "
for poly in polyList:
print str(poly) + '\n'
p1 = pointClass.Point(-122.3235, 47.603)
p2 = pointClass.Point(-122.321, 47.61)
p3 = pointClass.Point(-122.342878, 47.607403)
p4 = pointClass.Point(-122.326, 47.61131)
p5 = pointClass.Point(-122.322718, 47.605963)
p6 = pointClass.Point(-122.324000, 47.603995)
p7 = pointClass.Point(-122.328002, 47.609703)
p8 = pointClass.Point(-122.320331, 47.607742)
pList = [p1, p2, p3, p4, p5, p6, p7, p8]
for p in pList:
print "Point: " + str(p)
for ind in range(len(polyList)):
if polyList[ind].bPointInPoly(p):
print '\t' + "is in region number " + str(ind)
else:
print '\t' + "is out of region number " + str(ind)
def inROI(lat, lon, bc_lTargetPolygons):
recordPoint = pointClass.Point(lon, lat)
for site in bc_lTargetPolygons.value:
if site.bPointInPoly(recordPoint):
return True
return False