def test_save_and_load_work(self):
path = self.get_path('.shp.zip')
save(path, proj4LL, sourceGeometries)
targetProj4, targetGeometries = load(path)[:2]
self.assert_('+proj=longlat' in targetProj4)
for sourceGeometry, targetGeometry in zip(sourceGeometries, targetGeometries):
self.assert_(sourceGeometry.equals(targetGeometry))
def load(Class, path):
if not exists(path):
raise IOError
if path.endswith('.zip'):
import geometryIO
from crosscompute_table import pandas as pd
[
proj4,
geometries,
field_packs,
field_definitions,
] = geometryIO.load(path)
# Convert to (longitude, latitude)
normalize_geometry = geometryIO.get_transformGeometry(
proj4 or geometryIO.proj4LL)
geometries = [normalize_geometry(x) for x in geometries]
# Convert to (latitude, longitude)
geometries = transform_geometries(geometries, drop_z)
geometries = transform_geometries(geometries, flip_xy)
# Generate table
table = pd.DataFrame(
field_packs, columns=[x[0] for x in field_definitions])
table['WKT'] = [x.wkt for x in geometries]
else:
table = super(GeotableType, Class).load(path)
# TODO: Consider whether there is a better way to do this
table.interpret = create_bound_method(_interpret, table)
return table
def _assert_proj_match(self, shp, csv):
"""Ensure that the projections match before continuing"""
# Use fiona to open the shapefile as this includes the projection type
shapefile = fiona.open(shp)
# read the first line of the csv to get the projection
csv_proj = open(csv).readline()
# Parse the Projection to a dictionary
csv_proj = csv_proj.split('PROJ.4')[1]
pairs = [x.split('=') for x in csv_proj.split(' +') if x != '' and '=' in x]
csv_proj = {x[0]:x[1] for x in pairs}
# Iterate through and ensure all values match in both projection dicts
for key in csv_proj.keys():
if key in csv_proj and key in shapefile.crs:
try:
assert(str(csv_proj[key]) == str(shapefile.crs[key]))
except:
logger.error("csv and shp Projections Don't Match")
raise AssertionError("csv and shapefile Projections Don't Match")
# Save the state of the projection
# self.proj = shapefile.crs['proj']
# Determine the type of distance measure based on the input projections
# measure = 'euclidean' if self.proj == 'utm' else 'haversine'
from geometryIO import load
proj4 = load(shp)[0]
self.measure = 'haversine' if 'longlat' in proj4 else 'euclidean'
def load(Class, geotable_path):
lonlat_geometries = geometryIO.load(geotable_path,
targetProj4=geometryIO.proj4LL)[1]
lonlat_point = GeometryCollection(lonlat_geometries).centroid
longitude, latitude = lonlat_point.x, lonlat_point.y
zone_number, zone_letter = utm.from_latlon(latitude, longitude)[-2:]
return Class(zone_number, zone_letter)
def load(Class, source_path, utm_zone, defaults=None, alternates=None):
instances = []
if not defaults:
defaults = {}
if not alternates:
alternates = {}
geometries, field_packs, field_definitions = geometryIO.load(
source_path, targetProj4=utm_zone.proj4)[1:]
for index, (
geometry,
field_pack,
) in enumerate(zip(geometries, field_packs)):
d = {}
for field_value, (
field_name,
field_type,
) in zip(field_pack, field_definitions):
d[field_name] = field_value
instance = Class(id=d.pop('id', index),
geometry=geometry,
attributes=d)
for k, v in defaults.items():
setattr(instance, k, d.pop(k, d.pop(alternates.get(k), v)))
instances.append(instance)
return instances
def load_geotable(path):
if path.endswith('.csv'):
table = read_csv(path)
elif path.endswith('.zip'):
proj4, geometries, fields, definitions = geometryIO.load(path)
normalize_geometry = geometryIO.get_transformGeometry(proj4)
geometries = [normalize_geometry(x) for x in geometries]
table = DataFrame(fields, columns=[x[0] for x in definitions])
table['WKT'] = [x.wkt for x in geometries]
else:
raise UnsupportedFormat('cannot load geotable (%s)' % path)
return table
def test_save_and_load_attributes_work(self):
fieldPacks = [(
# 'Спасибо'.decode('utf-8'),
datetime.datetime(2000, 1, 1),
)]
fieldDefinitions = [
# ('String', OFTString),
('Date', OFTDate),
]
path = self.get_path()
save(path, proj4LL, sourceGeometries, fieldPacks, fieldDefinitions)
for sourceField, targetField in zip(fieldPacks[0], load(path)[2][0]):
self.assertEqual(sourceField, targetField)
def load_geotable(path):
if path.endswith('.csv'):
table = read_csv(path)
elif path.endswith('.zip'):
proj4, geometries, fields, definitions = geometryIO.load(path)
normalize_geometry = geometryIO.get_transformGeometry(proj4)
# Convert to (longitude, latitude)
geometries = [normalize_geometry(x) for x in geometries]
# Convert to (latitude, longitude)
geometries = transform_geometries(geometries, flip_xy)
table = DataFrame(fields, columns=[x[0] for x in definitions])
table['WKT'] = [x.wkt for x in geometries]
else:
raise UnsupportedFormat('cannot load geotable (%s)' % path)
return table
def __init__(self, shp, csv, **kwargs):
self.shp_p, self.csv_p = shp, csv
self.priority_metric = kwargs['prioritize'] if 'prioritize' in kwargs else 'population'
# logger.info('Asserting Input Projections Match')
# self._assert_proj_match(shp, csv)
from geometryIO import load
proj4 = load(shp)[0]
self.measure = 'haversine' if 'longlat' in proj4 else 'euclidean'
# Load in and align input data
logger.info('Aligning Network Nodes With Input Metrics')
self._network, self._metrics = prep_data( nx.read_shp(shp),
# pd.read_csv(csv, header=1),
pd.read_csv(csv),
loc_tol = self.TOL)
logger.info('Computing Pairwise Distances')
self.distance_matrix = self._distance_matrix()
if len(self.distance_matrix[(self.distance_matrix > 0) & (self.distance_matrix < self.TOL)]) > 0:
logger.error("""Dataset Contains Edges, Less Than {tolerance} Meters!
This can result in incorrect alignment of metrics and network,
where fake nodes are incorrectly assigned metrics.
This error is resolved by buffering your input data.""".format(tolerance=self.TOL))
# Set the edge weight to the distance between those nodes
self._weight_edges()
logger.info('Directing Network Away From Roots')
# Transform edges to a rooted graph
self.direct_network()
# Assert that the netork is a tree
self.assert_is_tree()
# Build list of fake nodes
self.fake_nodes = self.fakes(self.metrics.index)
#Fillna values with Zero
self._metrics = self.metrics.fillna(0)
def test_load_raises_exceptions(self):
# The format is unrecognized
path = self.get_path('')
with self.assertRaises(GeometryError):
load(path)
def test_load_with_targetProj4_works(self):
path = self.get_path()
save(path, proj4LL, sourceGeometries)
self.assert_('+proj=longlat' not in load(path, targetProj4=proj4SM)[0])
