def test_weighted_character_array(self):
area = self.df_buildings.geometry.area
sw = Queen_higher(k=3, geodataframe=self.df_tessellation)
weighted = mm.weighted_character(self.df_buildings,
self.df_tessellation, 'height', 'uID',
sw, area)
assert weighted[38] == 18.301521351817303
def test_mean_character(self):
self.df_tessellation['mesh'] = mm.mean_character(self.df_tessellation)
spatial_weights = Queen_higher(k=3, geodataframe=self.df_tessellation)
self.df_tessellation['mesh_sw'] = mm.mean_character(
self.df_tessellation, spatial_weights)
self.df_tessellation[
'area'] = area = self.df_tessellation.geometry.area
self.df_tessellation['mesh_ar'] = mm.mean_character(
self.df_tessellation, spatial_weights, values='area')
self.df_tessellation['mesh_array'] = mm.mean_character(
self.df_tessellation, spatial_weights, values=area)
self.df_tessellation['mesh_id'] = mm.mean_character(
self.df_tessellation, spatial_weights, values='area', rng=(10, 90))
self.df_tessellation['mesh_iq'] = mm.mean_character(
self.df_tessellation, spatial_weights, values='area', rng=(25, 75))
neighbours = spatial_weights.neighbors[38]
total_area = sum(self.df_tessellation.iloc[neighbours].geometry.area
) + self.df_tessellation.geometry.area[38]
check = total_area / (len(neighbours) + 1)
assert self.df_tessellation['mesh'][38] == check
assert self.df_tessellation['mesh_sw'][38] == check
assert self.df_tessellation['mesh_ar'][38] == check
assert self.df_tessellation['mesh_array'][38] == check
assert self.df_tessellation['mesh_id'][38] == 2206.611646069303
assert self.df_tessellation['mesh_iq'][38] == 2118.609142733066
gdf = self.df_tessellation
gdf.index = gdf.index + 20
with pytest.raises(ValueError):
self.df_tessellation['meshE'] = mm.mean_character(gdf)
def test_theil(self):
full = mm.theil(self.df_tessellation, 'area', order=3)
assert full[0] == 0.2574468431886534
sw = Queen_higher(k=3, geodataframe=self.df_tessellation)
full_sw = mm.theil(self.df_tessellation, 'area', spatial_weights=sw)
assert full_sw[0] == 0.2574468431886534
limit = mm.theil(self.df_tessellation,
'area',
spatial_weights=sw,
rng=(10, 90))
assert limit[0] == 0.13302952097969373
def test_simpson(self):
ht = mm.simpson(self.df_tessellation, 'area', order=3)
assert ht[0] == 0.385
sw = Queen_higher(k=3, geodataframe=self.df_tessellation)
ht_sw = mm.simpson(self.df_tessellation, 'area', spatial_weights=sw)
assert ht_sw[0] == 0.385
quan_sw = mm.simpson(self.df_tessellation,
'area',
spatial_weights=sw,
binning='quantiles',
k=3)
assert quan_sw[0] == 0.395
def test_rng(self):
full = mm.rng(self.df_tessellation, 'area', order=3)
assert full[0] == 8255.372874447059
sw = Queen_higher(k=3, geodataframe=self.df_tessellation)
full_sw = mm.rng(self.df_tessellation, 'area', spatial_weights=sw)
assert full_sw[0] == 8255.372874447059
area = self.df_tessellation['area']
full2 = mm.rng(self.df_tessellation, area, order=3)
assert full2[0] == 8255.372874447059
limit = mm.rng(self.df_tessellation,
'area',
spatial_weights=sw,
rng=(10, 90))
assert limit[0] == 4122.139212736442
def test_wall(self):
sw = Queen_higher(geodataframe=self.df_buildings, k=1)
wall = mm.wall(self.df_buildings)
wall_sw = mm.wall(self.df_buildings, sw)
assert wall[0] == wall_sw[0]
assert wall[0] == 137.2106961418436
def test_covered_area(self):
sw = Queen_higher(geodataframe=self.df_tessellation, k=1)
covered = mm.covered_area(self.df_tessellation)
covered_sw = mm.covered_area(self.df_tessellation, sw)
assert covered[0] == covered_sw[0]
assert covered[0] == 24115.66721833942
def test_weighted_character_sw(self):
sw = Queen_higher(k=3, geodataframe=self.df_tessellation)
weighted = mm.weighted_character(self.df_buildings,
self.df_tessellation, 'height', 'uID',
sw)
assert weighted[38] == 18.301521351817303
def simpson(objects, values, spatial_weights=None, order=None, binning='HeadTail_Breaks', **classification_kwds):
"""
Calculates the Simpson\'s diversity index of values within k steps of morphological tessellation
Uses `mapclassify.classifiers` under the hood for binning.
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing morphological tessellation
values : str, list, np.array, pd.Series
the name of the dataframe column, np.array, or pd.Series where is stored character value.
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix of set order will be calculated
based on objects.
order : int
order of Queen contiguity
binning : str
One of mapclassify classification schemes
Options are Box_Plot, Equal_Interval, Fisher_Jenks,
Fisher_Jenks_Sampled, HeadTail_Breaks, Jenks_Caspall,
Jenks_Caspall_Forced, Jenks_Caspall_Sampled, Max_P_Classifier,
Maximum_Breaks, Natural_Breaks, Quantiles, Percentiles, Std_Mean,
User_Defined
**classification_kwds : dict
Keyword arguments for classification scheme
For details see mapclassify documentation:
https://mapclassify.readthedocs.io/en/latest/api.html
Returns
-------
Series
Series containing resulting values.
References
----------
Examples
--------
"""
def simpson_di(data):
""" Given a hash { 'species': count } , returns the Simpson Diversity Index
>>> simpson_di({'a': 10, 'b': 20, 'c': 30,})
0.3888888888888889
https://gist.github.com/martinjc/f227b447791df8c90568
"""
def p(n, N):
""" Relative abundance """
if n == 0:
return 0
else:
return float(n) / N
N = sum(data.values())
return sum(p(n, N)**2 for n in data.values() if n != 0)
try:
import mapclassify.classifiers
except ImportError:
raise ImportError(
"The 'mapclassify' package is required to use the 'scheme' "
"keyword")
schemes = {}
for classifier in mapclassify.classifiers.CLASSIFIERS:
schemes[classifier.lower()] = getattr(mapclassify.classifiers,
classifier)
binning = binning.lower()
if binning not in schemes:
raise ValueError("Invalid binning. Binning must be in the"
" set: %r" % schemes.keys())
# define empty list for results
results_list = []
print('Calculating Simpson\'s diversity index...')
if spatial_weights is None:
print('Generating weights matrix (Queen) of {} topological steps...'.format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights = Queen_higher(k=order, geodataframe=objects)
else:
if not all(objects.index == range(len(objects))):
raise ValueError('Index is not consecutive range 0:x, spatial weights will not match objects.')
if values is not None:
if not isinstance(values, str):
objects['mm_v'] = values
values = 'mm_v'
bins = schemes[binning](objects[values], **classification_kwds).bins
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
neighbours = spatial_weights.neighbors[index]
values_list = objects.iloc[neighbours][values].tolist()
if values_list:
values_list.append(row[values])
else:
values_list = [row[values]]
sample_bins = mapclassify.classifiers.User_Defined(values_list, bins)
counts = dict(zip(bins, sample_bins.counts))
results_list.append(simpson_di(counts))
series = pd.Series(results_list)
if 'mm_v' in objects.columns:
objects.drop(columns=['mm_v'], inplace=True)
print('Simpson\'s diversity index calculated.')
return series
def theil(objects, values, spatial_weights=None, order=None, rng=(0, 100)):
"""
Calculates the Theil measure of inequality of values within k steps of morphological tessellation
Uses `pysal.explore.inequality.theil.Theil` under the hood.
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing morphological tessellation
values : str, list, np.array, pd.Series
the name of the dataframe column, np.array, or pd.Series where is stored character value.
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix of set order will be calculated
based on objects.
order : int
order of Queen contiguity
rng : Two-element sequence containing floats in range of [0,100], optional
Percentiles over which to compute the range. Each must be
between 0 and 100, inclusive. The order of the elements is not important.
Returns
-------
Series
Series containing resulting values.
References
----------
Examples
--------
"""
from pysal.explore.inequality.theil import Theil
# define empty list for results
results_list = []
print('Calculating Theil index...')
if spatial_weights is None:
print('Generating weights matrix (Queen) of {} topological steps...'.format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights = Queen_higher(k=order, geodataframe=objects)
else:
if not all(objects.index == range(len(objects))):
raise ValueError('Index is not consecutive range 0:x, spatial weights will not match objects.')
if values is not None:
if not isinstance(values, str):
objects['mm_v'] = values
values = 'mm_v'
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
neighbours = spatial_weights.neighbors[index]
values_list = objects.iloc[neighbours][values].tolist()
if values_list:
values_list.append(row[values])
else:
values_list = [row[values]]
if rng:
from momepy import limit_range
values_list = limit_range(values_list, rng=rng)
results_list.append(Theil(values_list).T)
series = pd.Series(results_list)
if 'mm_v' in objects.columns:
objects.drop(columns=['mm_v'], inplace=True)
print('Theil index calculated.')
return series
def node_density(objects,
nodes,
spatial_weights=None,
order=9,
node_id='nodeID'):
"""
Calculate the density of nodes within topological steps of morphological tessellation.
Node is marked as reached if reached cell con
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing tessellation objects to analyse
nodes : GeoDataFrame
GeoDataFrame containing nodes
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix of selected order will be calculated
based on objects.
order : int
order of Queen contiguity
node_id : str
name of the column of objects gdf with node id.
Returns
-------
Series
Series containing resulting values.
References
---------
Jacob
Notes
-----
"""
# define empty list for results
results_list = []
print('Calculating gross density...')
if not all(objects.index == range(len(objects))):
raise ValueError(
'Index is not consecutive range 0:x, spatial weights will not match objects.'
)
if spatial_weights is None:
print('Generating weights matrix (Queen) of {} topological steps...'.
format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights = Queen_higher(k=order, geodataframe=objects)
# iterating over rows one by one
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
neighbours = spatial_weights.neighbors[index]
neighbours.append(index)
sub = objects.iloc[neighbours]
subnodes = set(sub['nodeID'])
results_list.append((len(subnodes) / sum(sub.geometry.area)) * 10000)
series = pd.Series(results_list)
print('Gross density calculated.')
return series
def blocks_count(tessellation, block_id, spatial_weights=None, order=5):
"""
Calculates the weighted number of blocks
Number of blocks within `k` topological steps defined in spatial_weights weighted by the analysed area.
.. math::
\\frac{\\sum_{i=1}^{n} {blocks}}{\\sum_{i=1}^{n} area_{i}}
NOT SURE
Parameters
----------
tessellation : GeoDataFrame
GeoDataFrame containing morphological tessellation
block_id : str, list, np.array, pd.Series (default None)
the name of the objects dataframe column, np.array, or pd.Series where is stored block ID.
spatial_weights : libpysal.weights (default None)
spatial weights matrix - If None, Queen contiguity matrix of set order will be calculated
based on objects.
order : int (default 5)
order of Queen contiguity. Used only when spatial_weights=None.
Returns
-------
Series
Series containing resulting values.
References
----------
Jacob
Examples
--------
Notes
-----
Blocks count or blocks density?
"""
# define empty list for results
results_list = []
if not isinstance(block_id, str):
block_id['mm_bid'] = block_id
block_id = 'mm_bid'
if not all(tessellation.index == range(len(tessellation))):
raise ValueError(
'Index is not consecutive range 0:x, spatial weights will not match objects.'
)
if spatial_weights is None:
print('Generating weights matrix (Queen) of {} topological steps...'.
format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights = Queen_higher(k=order, geodataframe=tessellation)
print('Calculating blocks...')
for index, row in tqdm(tessellation.iterrows(),
total=tessellation.shape[0]):
neighbours = spatial_weights.neighbors[index]
neighbours.append(index)
vicinity = tessellation.iloc[neighbours]
results_list.append(
len(set(list(vicinity[block_id]))) / sum(vicinity.geometry.area))
series = pd.Series(results_list)
if 'mm_bid' in tessellation.columns:
tessellation.drop(columns=['mm_bid'], inplace=True)
print('Blocks calculated.')
return series
def gross_density(objects,
buildings,
area,
character,
spatial_weights=None,
order=3,
unique_id='uID'):
"""
Calculate the density
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing tessellation objects to analyse
buildings : GeoDataFrame
GeoDataFrame containing buildings
area : str
name of the column with area values
character : str
name of the column with values of target character for density calculation
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix of selected order will be calculated
based on objects.
order : int
order of Queen contiguity
unique_id : str
name of the column with unique id. If there is none, it could be generated by unique_id()
Returns
-------
Series
Series containing resulting values.
References
---------
Jacob??
Notes
-----
Rework, it is a mess. Terminology!
"""
# define empty list for results
results_list = []
print('Calculating gross density...')
if not all(objects.index == range(len(objects))):
raise ValueError(
'Index is not consecutive range 0:x, spatial weights will not match objects.'
)
if spatial_weights is None:
print('Generating weights matrix (Queen) of {} topological steps...'.
format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights = Queen_higher(k=order, geodataframe=objects)
# iterating over rows one by one
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
neighbours_id = spatial_weights.neighbors[index]
neighbours_id.append(index)
neighbours = objects.iloc[neighbours_id]
fa = buildings.loc[buildings[unique_id].isin(
neighbours[unique_id])][character]
results_list.append(sum(fa) / sum(neighbours[area]))
series = pd.Series(results_list)
print('Gross density calculated.')
return series
def building_adjacency(objects,
tessellation,
spatial_weights=None,
spatial_weights_higher=None,
order=3,
unique_id='uID'):
"""
Calculate the level of building adjacency
Building adjacency reflects how much buildings tend to join together into larger structures.
It is calculated as a ratio of joined built-up structures and buildings within k topological steps.
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing objects to analyse
tessellation : GeoDataFrame
GeoDataFrame containing morphological tessellation - source of spatial_weights and spatial_weights_higher.
It is crucial to use exactly same input as was used durign the calculation of weights matrix and spatial_weights_higher.
If spatial_weights or spatial_weights_higher is None, tessellation is used to calulate it.
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix will be calculated
based on tessellation
spatial_weights_higher : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity of higher order will be calculated
based on tessellation
order : int
Order of Queen contiguity
Returns
-------
Series
Series containing resulting values.
References
---------
Vanderhaegen S and Canters F (2017) Mapping urban form and function at city
block level using spatial metrics. Landscape and Urban Planning 167: 399–409.
Examples
--------
>>> buildings_df['adjacency'] = momepy.building_adjacency(buildings_df, tessellation_df, unique_id='uID')
Calculating adjacency...
Calculating spatial weights...
Spatial weights ready...
Generating weights matrix (Queen) of 3 topological steps...
Generating dictionary of built-up patches...
100%|██████████| 144/144 [00:00<00:00, 9301.73it/s]
Calculating adjacency within k steps...
100%|██████████| 144/144 [00:00<00:00, 335.55it/s]
Adjacency calculated.
>>> buildings_df['adjacency'][10]
0.23809523809523808
"""
# define empty list for results
results_list = []
print('Calculating adjacency...')
if not all(tessellation.index == range(len(tessellation))):
raise ValueError(
'Index is not consecutive range 0:x, spatial weights will not match objects.'
)
# if weights matrix is not passed, generate it from objects
if spatial_weights is None:
print('Calculating spatial weights...')
from libpysal.weights import Queen
spatial_weights = Queen.from_dataframe(objects, silence_warnings=True)
print('Spatial weights ready...')
if spatial_weights_higher is None:
print('Generating weights matrix (Queen) of {} topological steps...'.
format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights_higher = Queen_higher(k=order,
geodataframe=tessellation)
print('Generating dictionary of built-up patches...')
# dict to store nr of courtyards for each uID
patches = {}
jID = 1
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
# if the id is already present in courtyards, continue (avoid repetition)
if index in patches:
continue
else:
to_join = [index
] # list of indices which should be joined together
neighbours = [] # list of neighbours
weights = spatial_weights.neighbors[
index] # neighbours from spatial weights
for w in weights:
neighbours.append(w) # make a list from weigths
for n in neighbours:
while n not in to_join: # until there is some neighbour which is not in to_join
to_join.append(n)
weights = spatial_weights.neighbors[n]
for w in weights:
neighbours.append(
w
) # extend neighbours by neighbours of neighbours :)
for b in to_join:
patches[b] = jID # fill dict with values
jID = jID + 1
print('Calculating adjacency within k steps...')
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
uid = tessellation.loc[tessellation[unique_id] ==
row[unique_id]].index[0]
neighbours = spatial_weights_higher.neighbors[uid]
neighbours_ids = tessellation.iloc[neighbours][unique_id]
neighbours_ids = neighbours_ids.append(
pd.Series(row[unique_id], index=[index]))
building_neighbours = objects.loc[objects[unique_id].isin(
neighbours_ids)]
indices = list(building_neighbours.index)
patches_sub = [patches[x] for x in indices]
patches_nr = len(set(patches_sub))
results_list.append(patches_nr / len(building_neighbours))
series = pd.Series(results_list)
print('Adjacency calculated.')
return series
def mean_interbuilding_distance(objects,
tessellation,
unique_id,
spatial_weights=None,
spatial_weights_higher=None,
order=3):
"""
Calculate the mean interbuilding distance within x topological steps
Interbuilding distances are calculated between buildings on adjacent cells based on `spatial_weights`.
.. math::
Parameters
----------
objects : GeoDataFrame
GeoDataFrame containing objects to analyse
tessellation : GeoDataFrame
GeoDataFrame containing morphological tessellation - source of spatial_weights and spatial_weights_higher.
It is crucial to use exactly same input as was used durign the calculation of weights matrix and spatial_weights_higher.
If spatial_weights or spatial_weights_higher is None, tessellation is used to calulate it.
unique_id : str
name of the column with unique id
spatial_weights : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity matrix will be calculated
based on tessellation
spatial_weights_higher : libpysal.weights, optional
spatial weights matrix - If None, Queen contiguity of higher order will be calculated
based on tessellation
order : int
Order of Queen contiguity
Returns
-------
Series
Series containing resulting values.
References
---------
ADD, but it is adapted quite a lot.
Notes
-----
Fix terminology, it is unclear.
Fix UserWarning.
Examples
--------
>>> buildings_df['mean_interbuilding_distance'] = momepy.mean_interbuilding_distance(buildings_df, tessellation_df, 'uID')
Calculating mean interbuilding distances...
Generating weights matrix (Queen)...
Generating weights matrix (Queen) of 3 topological steps...
Generating adjacency matrix based on weights matrix...
Computing interbuilding distances...
100%|██████████| 746/746 [00:03<00:00, 200.14it/s]
Computing mean interbuilding distances...
100%|██████████| 144/144 [00:00<00:00, 317.42it/s]
Mean interbuilding distances calculated.
>>> buildings_df['mean_interbuilding_distance'][0]
29.305457092042744
"""
if not all(tessellation.index == range(len(tessellation))):
raise ValueError(
'Index is not consecutive range 0:x, spatial weights will not match objects.'
)
print('Calculating mean interbuilding distances...')
if spatial_weights is None:
print('Generating weights matrix (Queen)...')
from libpysal.weights import Queen
# matrix to capture interbuilding relationship
spatial_weights = Queen.from_dataframe(tessellation)
if spatial_weights_higher is None:
print('Generating weights matrix (Queen) of {} topological steps...'.
format(order))
from momepy import Queen_higher
# matrix to define area of analysis (more steps)
spatial_weights_higher = Queen_higher(k=order,
geodataframe=tessellation)
# define empty list for results
results_list = []
print('Generating adjacency matrix based on weights matrix...')
# define adjacency list from lipysal
adj_list = spatial_weights.to_adjlist()
adj_list['distance'] = -1
print('Computing interbuilding distances...')
# measure each interbuilding distance of neighbours and save them to adjacency list
for index, row in tqdm(adj_list.iterrows(), total=adj_list.shape[0]):
inverted = adj_list[(adj_list.focal == row.neighbor)][(
adj_list.neighbor == row.focal)].iloc[0]['distance']
if inverted == -1:
object_id = tessellation.iloc[row.focal.astype(int)][unique_id]
building_object = objects.loc[objects[unique_id] == object_id]
neighbours_id = tessellation.iloc[row.neighbor.astype(
int)][unique_id]
building_neighbour = objects.loc[objects[unique_id] ==
neighbours_id]
adj_list.loc[
index,
'distance'] = building_neighbour.iloc[0].geometry.distance(
building_object.iloc[0].geometry)
else:
adj_list.at[index, 'distance'] = inverted
print('Computing mean interbuilding distances...')
# iterate over objects to get the final values
for index, row in tqdm(objects.iterrows(), total=objects.shape[0]):
# id to match spatial weights
uid = tessellation.loc[tessellation[unique_id] ==
row[unique_id]].index[0]
# define neighbours based on weights matrix defining analysis area
neighbours = spatial_weights_higher.neighbors[uid]
neighbours.append(uid)
if neighbours:
selection = adj_list[adj_list.focal.isin(neighbours)][
adj_list.neighbor.isin(neighbours)]
results_list.append(np.nanmean(selection.distance))
series = pd.Series(results_list)
print('Mean interbuilding distances calculated.')
return series