Searoute py
A python package for generating the shortest sea route between two points on Earth.
If points are on land, the function will attempt to find the nearest point on the sea and calculate the route from there.
Not for routing purposes! This library was developed to generate realistic-looking sea routes for visualizations of maritime routes, not for mariners to route their ships.
Installation
pip install searoute
Usage
import searoute as sr
origin = [0.3515625, 50.064191736659104]
destination = [117.42187500000001, 39.36827914916014]
route = sr.searoute(origin, destination)
print("{:.1f} {}".format(route.properties['length'], route.properties['units']))
routeMiles = sr.searoute(origin, destination, units="mi")
Bring your network :
my_nodes = {
(1, 2): {'x': 1, 'y': 2},
(2, 2): {'x': 2, 'y': 2}
}
my_edges = {
(1, 2): {
(2, 2): {"weight": 10, "other_attr": "some_value"}
}
}
myM = sr.from_nodes_edges_set(sr.Marnet(), my_nodes, my_edges)
myP = sr.from_nodes_edges_set(sr.Ports(), my_nodes, None)
route_with_my_ports = sr.searoute(origin, destination, P = myP, include_ports=True)
route_with_my_ntw = sr.searoute(origin, destination, P = myP, M = myM )
Nodes and Edges
Nodes
A node (or vertex) is a fundamental unit of which the graphs Ports and Marnet are formed.
In searoute, a node is represented by it's id as a tuple
of lon,lat, and it's attributes:
x
: float ; requiredy
: float ; requiredt
: bool ; optional. default is False
, will be used for filtering ports with terminals
.
{
(lon:float, lat:float): {'x': lon:float, 'y': lat:float, *args:any},
...
}
Edges
An edge is a line or connection between a note and an other. This connection can be represented by a set of node_id->node_id with it's attributes:
weight
: float ; required. Can be distance, cost etc... if not set will by default calculate the distance between nodes using Haversine formula.passage
: str ; optional. Can be one of Passage (searoute.classes.passages.Passage). If not not set, no restriction will be applied. If set make sure to update Marnet().restrictions = [...]
with your list of passages to avoid.
{
(lon:float, lat:float)->node_id: {
(lon:float, lat:float)->node_id: {
"weight": distance:float,
*args: any}
},
...
}
Example with more parameters :
route = sr.searoute(origin, destination, append_orig_dest=True, restrictions=['northwest'], include_ports=True, port_params={'only_terminals':True, 'country_pol': 'FR', 'country_pod' :'CN'})
returns :
{
"geometry": {
"coordinates": [],
"type": "LineString"
},
"properties": {
"duration_hours": 461.88474412693756,
"length": 20529.85310695412,
"port_origin": {
"cty": "France",
"name": "Le Havre",
"port": "FRLEH",
"t": 1,
"x": 0.107054,
"y": 49.485998
},
"port_dest": {
"cty": "China",
"name": "Tianjin",
"port": "CNTSN",
"t": 1,
"x": 117.744852,
"y": 38.986802
},
"units": "km"
},
"type": "Feature"
}
Preferred Ports
It's possible to select referred ports which can be configured with one or a list
of AreaFeature
:
Start by creating/referencing preferred ports using PortProps
object which should contain a port_id
, share
(could be any positive number) and props
corresponding to attributes of a port.
port_one = PortProps('MY_PORT_ID', 2, {'x':1, 'y':2})
port_two = PortProps('USNYC', 1)
Initiate a AreaFeature
with its coords and a name, as well as list
of preferred_ports
(could be a list
, str
, or PortProps
)
area_one = AreaFeature(coordinates=[[[0,0], [0,10],[0,20], [20, 20], [0, 0]]], name= 'Special_Name', preferred_ports=[port_one, port_two])
area_two = AreaFeature(...)
Note that the smallest AreaFeature which contains the point will be selected.
Finally, call the function which will return a tuple of 3 values, or 4 values when include_area_name
is set to True
:
origin = (11, 12)
pref_ports = myPorts.get_preferred_ports(*origin, AreaFeature.create([area_one, area_two]), top=2, include_area_name = True)
Usage in main function
areas = AreaFeature.create([area_one, area_two])
sr.searoute(..., include_ports = True, port_params = {'ports_in_areas': areas})
Parameters
origin
Mandatory. A tuple or array of 2 floats representing longitude and latitude i.e : ({lon}, {lat})
destination
Mandatory. A tuple or array of 2 floats representing longitude and latitude i.e : ({lon}, {lat})
units
Optional. Default to km
= kilometers, can be m
= meters mi
= miles ft
= feets in
= inches deg
= degrees cen
= centimeters rad
= radians naut
= nauticals yd
= yards
speed_knot
Optional. Speed of the boat, default 24 knots
append_orig_dest
Optional. If the origin and destination should be appended to the LineString, default is False
restrictions
Optional. List of passages to be restricted during calculations.
Possible values : babalmandab
, bosporus
, gibraltar
, suez
, panama
, ormuz
, northwest
, malacca
, sunda
, chili
, south_africa
;
default is ['northwest']
include_ports
Optional. If the port of load and discharge should be included, default is False
port_params
Optional. If include_ports
is True
then you can set port configuration for advanced filtering :
only_terminals
to include only terminal ports, default False
country_pol
country iso code (2 letters) for closest port of load, default None
. When not set or not found, closest port is based on geography.country_pod
country iso code (2 letters) for closest port of discharge, default None
. When not set or not found, closest port is based on geography.country_restricted
to filter out ports that have registered an argument key to_cty
(a list) which indicates an existing route to the country same as in country_pod
; default False
ports_in_areas
: a FeatureCollection containing areas with preferred ports, created of AreaFeature, use AreaFeature.create([...]). The previous configurations will be ignored.
If there are many ports then the result will be a list of GeoJson Features, instead of an object of GeoJson Feature.
Preferred ports with share = 0 will be ignored.
return_passages
Optional. to return traversed passages, default is False
default is {}
Returns
GeoJson Feature or list[GeoJson Feature] if many routes configured in port_params
.
Credits
- NetworkX, a Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks.
- GeoJson, a python package for GeoJSON
- turfpy, a Python library for performing geo-spatial data analysis which reimplements turf.js. (up to version
searoute 1.1.0
) - OSMnx, for geo-spacial networks. (up to version
searoute 1.1.0
) - Eurostat's Searoute Java library
Change Log
1.0.2 (30/09/2022)
- First Release, published module
1.0.3 (04/10/2022)
- changed the core of the calculation to osmnx library for better performance
- making 1.0.2 obsolete
- added duration measurement together with length in the LineString properties
1.0.4 (04/10/2022)
- fix requirements of the module
1.0.5 (04/10/2022)
- Adjusted network (.json) by connecting all LineStrings in order to return a route from any location to any location
1.0.6 (05/10/2022)
- Improved network connectivity
1.0.7 (13/10/2022)
- Improved network connectivity
- Connection points were not adjusted
1.0.8 (25/10/2022)
- Improved network connectivity
- Added append_orig_dest parameter as input to append origin and destination points to the LineString
- Updated README.md
1.0.9 (04/01/2023)
- Improved network connectivity
- Added restrictions parameter as input to control restrictiveness of connections due to different reasons
- Updated README.md
1.1.0 (23/05/2023)
- Improved network connectivity
- Added new feature : closest port of load and port of discharge
- Updated README.md
1.2.0 (24/08/2023)
- 35x Faster: Significantly accelerated compared to v1.1 by rethinking and reorganizing the code.
- Re-think Class Design: Re-structured and simplified class structure.
- Easy API Access
- Improved Network Connectivity: Enhanced network reliability.
- Updated README.md: Revised for the latest information.
1.2.1 (28/08/2023)
- Patch fixed issue when querying country with no port result
1.2.2 (08/09/2023)
- Patch fixed and simplified coords line normalization
1.2.3 (05/01/2024)
- Patch fixed and improved coords validation
1.3.0 (11/01/2024)
- Added in port_params additional parameter country_restricted
- Updated license to Apache 2.0
1.3.1 (27/01/2024)
- Enriched network with additional passages
1.4.0 (09/06/2024)
- Enriched network with additional passages : chili and south_africa (#27)
- Adjusted logic of port selection (#28)
- Add feature to suggest preferred port of load and port of discharge (#29)
- Fixed del property
to_cty
in main function (#30) - Created unit tests
- return passages traversed by the path (#34)
1.4.1 (22/06/2024)
- Fixed support python 3.7 3.8 (issue #36)
- Fixed bug in
return_passages
(issue #35) - Added tests for
return_passages
1.4.2 (22/08/2024)
- Improved Marnet around Baltic sea #37
- Fixed issue with returned passage when crossing the ocean #35
- Adding unit tests
- Added test for mandatory passages example