add setup.py

• example --> pysal/spaghetti

• regular lattice
• "waverly"

• v0.0-prealpha
• v0.1-prealpha
• v0.1-alpha
• v0.1-beta
• v0.1-rc
• v0.1
• release on PyPI and conda-forge

Add edge test cases:

• utils._drop_geoms() 640d661
• utils._weld_MultiLineString()
  • 3+ parts with one set of equal sp/ep removed code; unnecessary in workflow
  • sp1 almost equals ep2 016b775
• utils.ring_correction()
  • disconnected ring 6c05b87
  • overlapping ring 6c05b87
• utils.record_filter()
  • column index ed038e7
  • single value in column ed038e7
  • multi value in column ed038e7
• utils.restriction_welder()
  • None-type KeyError 4b455ba
  • welding MultiLineString 81c685a
• utils.line_splitter()
  • no __iter__ attribute 4136694 refactored code
• utils._create_split_lines()
  • unaltered line 7d67f19
  • create endpoint for ring roads
• utils._find_break_locs()
  • case 1 56a9600
  • case 2 56a9600
  • case 3 68d129f
  • case 4 779ef85
  • case 5 (single) 779ef85
  • case 5 (multi) b8280d6
• utils._make_break_locs()
  • breaks and standard 56a9600
  • breaks and not standard line/mline 779ef85

• add PyPI version badge
• rearrange badges/location in README.md

See this issue in snkit for description.

add Quickstart following the example from pysal/spaghetti

set up Travis CI

• which Python versions to officially support?
• .travis.yml — spaghetti example
• .coveragerc — spaghetti example

Add momepy as a related project in README.md. @martinfleis would you be able to in a PR for this to test contributors in release note generation?

Update labels for notes customization and try PR with more than one label (e.g. enhancement and tiger_edges).

See here.

Add related project list to README.md.

BY far the two largest time/memory hogs are [n2n_matrix] and [paths] within Network.cost_matrix():

# calculate shortest path length and records paths if desired
n2n_matrix, paths = utils.shortest_path(self, gp=wpaths)

An attempt to speed up should be made...

• try numba?
• simple multiprocessing?

To Do's for README.md:

• Title/Documentation Link
• add functionality blurb — #4
• badges
  • GitHub tag badge — 6a863bc
  • travis build badge
  • python versions badge
  • DOI badge
  • coverage
  • code style (black)
  • anaconda stuff
• Installation
• Requirements
• License
• DOI badge
• BibTeX Citation

Add tests (for both synthetic and empirical):

• network diameter
• network radius
• circuity

Generate a conda-forge/staged-recipe.

add a .gitignore file — spaghetti example

add a documentation website similar to spaghetti

• from jGaboardi_dissertation structure (at1866_Master):

  • tigernet/__init__.py
  • tigernet/spaghetti.py
  • tigernet/sauce.py
  • tigernet/utils.py
  • tigernet/dissertation_workflows.py

• to jGaboardi/tigernet structure:

  • tigernet/__init__.py
  • tigernet/spaghetti.py --> tigernet/tigernet.py
  • tigernet/sauce.py
    • tigernet/dataprep.py, tigernet/netbuild.py, tigernet/point_funcs.py
  • tigernet/utils.py
    • tigernet/utils.py (utils+sauce+dissertation_workflows)?
    • integrate with others?
  • tigernet/dissertation_workflows.py

Following #61

• add auto merge bot to tigernet-feedstock
• add conda-forge badge to README
• add recipe badge to README
• sync description here with that of tigernet-feedstock

Currently, a sizable portion of the CI tests are disabled when running Windows due to an undiagnosed issue resulting in incorrect results. Periodically go back and check wether these tests are passing or not.

• short blurb about functionality on README.md
• see example in pysal/spaghetti

Probably should convert all lookups to dictionaries from lists. Not sure why I made all these lists in dissertation implementation...

See also #29

To reproduce:

bbox = (-84.279,30.480,-84.245,30.505)
f = "zip://test_data/Edges_Leon_FL_2010.zip!Edges_Leon_FL_2010.shp"
gdf = geopandas.read_file(f, bbox=bbox)
gdf = gdf.to_crs("epsg:2779")
yes_roads = gdf["MTFCC"].str.startswith("S")
roads = gdf[yes_roads].copy()

kwargs = {"s_data": roads.copy(), "from_raw": True}
attr_kws = {"attr1": ATTR1, "attr2": ATTR2}
kwargs.update(attr_kws)
comp_kws = {"record_components": True, "largest_component":True}
kwargs.update(comp_kws)
graph_elems_kws = {"def_graph_elems": True}
kwargs.update(graph_elems_kws)
geom_kws = {"record_geom": True, "calc_len": True}
kwargs.update(geom_kws)
mtfcc_kws = {"discard_segs": discard_segs, "skip_restr": SKIP_RESTR}
mtfcc_kws.update({"mtfcc_split": INTRST, "mtfcc_intrst": INTRST})
mtfcc_kws.update({"mtfcc_split_grp": SPLIT_GRP, "mtfcc_ramp": RAMP})
mtfcc_kws.update({"mtfcc_split_by": SPLIT_BY, "mtfcc_serv": SERV_DR})
kwargs.update(mtfcc_kws)
net = tigernet.Network(**kwargs)
net.calc_net_stats()

IndexError                                Traceback (most recent call last)
<ipython-input-84-bab5579caf99> in <module>
     14 kwargs.update(mtfcc_kws)
     15 net = tigernet.Network(**kwargs)
---> 16 net.calc_net_stats()

~/tigernet/tigernet/tigernet.py in calc_net_stats(self, conn_stat)
    622 
    623         # Calculate the sinuosity of network segments and provide descriptive stats
--> 624         stats.calc_sinuosity(self)
    625 
    626         # Set node degree attributes

~/tigernet/tigernet/stats.py in calc_sinuosity(net)
     18     # Calculate absolute shortest path along segments
     19     # Euclidean distance from vertex1 to vertex2
---> 20     net.s_data = utils.euc_calc(net, col="euclid")
     21 
     22     # Calculate sinuosity for segments

~/tigernet/tigernet/utils.py in euc_calc(net, col)
   1274     net.s_data[col] = numpy.nan
   1275     for (seg_k, (n1, n2)) in net.segm2node:
-> 1276         p1, p2 = net.node2coords[n1][1][0], net.node2coords[n2][1][0]
   1277         ed = _euc_dist(p1, p2)
   1278         net.s_data.loc[(net.s_data[net.sid_name] == seg_k), col] = ed

IndexError: list index out of range

create DOI through zenodo

add MANIFEST.in in top directory

• example in pysal/spaghetti

Current failures Python 3.9 support:

======================================================================= short test summary info =======================================================================
FAILED tigernet/tests/test_tigernet_empirical_gdf.py::TestNeworkTopologyEmpiricalGDF::test_network_node2node - AssertionError: Lists differ: [5, 8, 9] != [8, 9, 5]
FAILED tigernet/tests/test_tigernet_empirical_gdf.py::TestNeworkComponentsEmpiricalGDF::test_network_ndata_components - AssertionError: Lists differ: [360, 99, 99, ...
FAILED tigernet/tests/test_tigernet_empirical_gdf.py::TestNeworkComponentsEmpiricalGDF::test_network_node_components_largest - KeyError: 281
FAILED tigernet/tests/test_tigernet_empirical_gdf.py::TestNetworkSimplifyEmpiricalGDF::test_simplify_copy_node_cc - KeyError: 89
FAILED tigernet/tests/test_tigernet_empirical_gdf.py::TestNetworkSimplifyEmpiricalGDF::test_simplify_inplace_node_cc - KeyError: 89
FAILED tigernet/tests/test_tigernet_synthetic.py::TestNetworkSimplifyBarb::test_simplify_copy_segm_cc - AssertionError: {2: [0, 1, 2]} != {1: [0, 1, 2]}
FAILED tigernet/tests/test_tigernet_synthetic.py::TestNetworkSimplifyBarb::test_simplify_inplace_segm_cc - AssertionError: {2: [0, 1, 2]} != {1: [0, 1, 2]}

See first attempt in #37.

Remove functionality to read in .shp files. Only relying on geopandas.GeoDataFrame as input is acceptable and more robust.

There appears to be a major bug that has cropped up since the dissertation version on this code.

• It seems to be within utils.update_adj(), but can't pinpoint where exactly it happens.
• It occurs when def_graph_elems=True

To reproduce:

bbox = (-84.279,30.480,-84.245,30.505)
f = "zip://test_data/Edges_Leon_FL_2010.zip!Edges_Leon_FL_2010.shp"
gdf = geopandas.read_file(f, bbox=bbox)
gdf = gdf.to_crs("epsg:2779")
yes_roads = gdf["MTFCC"].str.startswith("S")
roads = gdf[yes_roads].copy()

kwargs = {"s_data": roads.copy(), "from_raw": True}
attr_kws = {"attr1": ATTR1, "attr2": ATTR2}
kwargs.update(attr_kws)
comp_kws = {"record_components": True, "largest_component":True}
kwargs.update(comp_kws)
graph_elems_kws = {"def_graph_elems": True}
kwargs.update(graph_elems_kws)
geom_kws = {"record_geom": True, "calc_len": True}
kwargs.update(geom_kws)
mtfcc_kws = {"discard_segs": discard_segs, "skip_restr": SKIP_RESTR}
mtfcc_kws.update({"mtfcc_split": INTRST, "mtfcc_intrst": INTRST})
mtfcc_kws.update({"mtfcc_split_grp": SPLIT_GRP, "mtfcc_ramp": RAMP})
mtfcc_kws.update({"mtfcc_split_by": SPLIT_BY, "mtfcc_serv": SERV_DR})
kwargs.update(mtfcc_kws)
net = tigernet.Network(**kwargs)

IndexError: list index out of range

add setup.py

Debug unittest failures that started here. At first glance this appears it could be due to the release of numpy v1.20.1.

To Do:

• start by pinning numpy v1.20.0
• actually pyproj issue (#67)

After some consideration, this repo will serve as a stub for the tigernet implementation developed for Gaboardi (2019), which can be cited in future publications through its DOI. Currently, some of the concepts are already being incorporated into spaghetti, with more of the functionality in the original tigernet potential (such as network measures pysal/spaghetti#126).

• Gaboardi, James D. 2019. Populated Polygons to Networks: A Population-Centric Approach to Spatial Network Allocation. ProQuest Dissertations Publishing.

Something changed between pyproj 2.6.1.post1 and pyproj 3.0.0.post1 that is leading to small variations in projected coordinates, which is then propagated on to point snapping distance, cost matrices, etc. For now, pinning to 2.6.1.post1 is reasonable (#66). A longer term solution is either updating the tests to reflect the altered projection results or figuring out if this is an actual bug. The code chunk below demonstrates the output difference between pyproj 2.6.1.post1 (top) and pyproj 3.0.0.post1 (bottom) with an example through geopandas and a MWE with only pyproj.

@knaaptime @slumnitz @martinfleis Have any of you noticed this in your work?

import warnings
warnings.filterwarnings("ignore")
import geopandas
import numpy
import pyproj
import shapely

packages = [geopandas, numpy, pyproj, shapely]
for p in packages:
    print(f"{p.__name__}: {p.__version__}")

layer = "WeightedParcels_Leon_FL_2010"
fin = f"zip://{layer}.zip!{layer}.shp"
bbox = (-84.279, 30.480, -84.245, 30.505)
gdf = geopandas.read_file(fin, bbox=bbox)
TEST_PARCEL = "1117160000020"
gdf = gdf[gdf["PARCEL_ID"] == TEST_PARCEL].reset_index(drop=True)
point = gdf.geometry[0]
x, y = point.x, point.y

from_crs, to_crs = gdf.crs, "EPSG:2779"

print(f" -- {geopandas.__name__}/{shapely.__name__} --")
print(gdf.crs, point)
gdf = gdf.to_crs(to_crs)
point = gdf.geometry[0]
print(gdf.crs, point)

print(f" -- {pyproj.__name__} --")
transformer = pyproj.Transformer.from_crs(from_crs, to_crs)
print(from_crs, (x, y))
print(to_crs, transformer.transform(y, x))

geopandas: 0.8.1
numpy: 1.19.4
pyproj: 2.6.1.post1
shapely: 1.7.1
 -- geopandas/shapely --
epsg:4326 POINT (-84.25873245332716 30.48436516338524)
EPSG:2779 POINT (623164.270749338 164564.2000569425)
 -- pyproj --
epsg:4326 (-84.25873245332716, 30.48436516338524)
EPSG:2779 (623164.270749338, 164564.20005694253)

geopandas: 0.8.2
numpy: 1.20.1
pyproj: 3.0.0.post1
shapely: 1.7.1
 -- geopandas/shapely --
epsg:4326 POINT (-84.25873245332716 30.48436516338524)
EPSG:2779 POINT (623164.6468275142 164563.5686168194)
 -- pyproj --
epsg:4326 (-84.25873245332716, 30.48436516338524)
EPSG:2779 (623164.6468275142, 164563.56861681942)

Remove all reference and code dealing with TIGER/Line Roads. Only using TIGER/Line Edges now.

add requirements.txt

reiterate need to wrap this up for publication.

use nose as the testing structure.

• example with pysal/spaghetti tests

• adopt black as the standard code format

• add badge

• use commit hook? --> cenpy example