-
Notifications
You must be signed in to change notification settings - Fork 9
/
todo.txt
77 lines (73 loc) · 3.49 KB
/
todo.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
MUST
OPTIONAL
measure gt-abstract avg degree growth
make visual distinction between covered and reached pois in cover plots
add comments to all functions, https://www.datacamp.com/community/tutorials/docstrings-python
simulate random link growth?
split up functions.py?
existing analysis: union carall and biketrack for biketrackcarall?
zip result pickles?
profile code?
DONE
X make coverage + lcchighlight plot for paper (milan?, london?, moscow?)
X fix: analysisplots: pois covered railwaystation not grid
X plots: for each city: length ratio of grown/existing versus ratio of metric (efficiency)
X calculate efficiencies for carminusbike simplified graphs - sayat
X check+fix GT chicago: lat lon mixed up for length calculations?!
X calculate efficiencies for simplified graphs
X clean up analysis plots
X simulate random GT growth
X implement pop density pois? <- no reliable data
X remove bicycle~designated and re-run where relevant (e.g. portland, chicago)
X use shapely simplify (dont use - covers become too imprecise)
X external plot and results folder due to github size limites https://stackoverflow.com/questions/38768454/repository-size-limits-for-github-com
X make test run of everything on a testcity
X add living_street and cyclestreet to biketrack, for netherlands
X get missing shape files
X location MultiPolygon: get largest polygon for cases like Amsterdam, account for Tokyo
X location polygon: fill holes
X styling for analysis plots
X osmnx 0.16 compatibility updates: geocode_to_gdf, etc
X add largest US/UK cities
X snap POIs to simplfied networks
X Export carconstrictedbike networks for betweenness calculations
X investigate/fix components measures for carminusbike network
X cover plots
X auto adjust thickness for gt abstract networks
X investigate/fix overlap measures
X remove cluster algorithms
X polish plots
X instead of turning streets completely into bike lanes, decrease speed (=increase effective distance) from 50km/h to 5km/h
X implement directness_lcc, length_lcc
X analysis-plot: add levels for: existing net (biketrack, and bikeable)
X analysis-plot: add overlap with existing net
X measure overlap
X measure relative cost and efficiency (to MST)
X analyze and plot metrics also for car setminus created bike networks
X fix delete_overlaps
X calculate metrics in a smart, additive way, especially coverage
X fix efficiency_local 0
X fix poi_coverage 0
X fix non-positive weight errors that happen in 03,04 for some cities
X fix division by zero in 05, in calculate_directness: return total_distance_haversine / total_distance_network
X pois should be created from data network file, not from online OSMNX. otherwise online changes will make data inconsistent if 01 and 02 are run at different times.
X re-run 03 for poi_source grid
X fix all *_poi_grid_nnidscarall.csv being empty!
X compress data to be commitable (below 100MB)
X loop 06 through cities
X consider edge cases: <2 POIs, <2 clusters, etc
X fix speed issue with buffer
X measure efficiency of existing infrastructure (simplified graphs, subsample)
X calculate coverage along edges, not nodes
X pull constants and parameters
X measure local/global efficiency of abstract GT
X analyze cluster data
X implement additional absolute cluster cutoff
X fix and test grid generation with numpy
X analysis in placeid loop
X make 01 more elegant (dict of Gs, involve all osmnxparameters)
X move plots into 06
X bundle all code together in a for loop over placeids
X delete unnecessary imports
X save cluster data
X create biketrackcarall networks in advance, and test