a node.js library for performing geospatial operations with geojson
All features are written in a functional manner with no side effects. In nearly all cases, they accept objects created by the point, linestring, polygon, and featurecollection functions, but these are simply for convenience. Any valid geojson Feature of FeatureCollection will do.
npm install turfTurf can also be run in a browser. To use it, download the minified file, and include it in a script tag.
<script src="turf.min.js"></script> note: This module is under active development and is in a pre-release form. The first official release is planned mid November 2013. Most features are pretty stable, but expect some changes periodically up until then.
Features
- load
- save
- point
- linestring
- polygon
- featurecollection
- extent
- square
- center
- bboxPolygon
- envelope
- centroid
- explode
- combine
- distance
- buffer
- nearest
- tin
- grid
- planepoint
- inside
- midpoint
- quantile
- jenks
- contour
Planned Features
Additional feature requests welcomed and encouraged. To request a feature, please add a github issue with a description.
- bezier
- interpolate
- tag
- area
- filter
- intersect
- reclass
- remove
- union
- erase
- smooth
Examples:
load
Loads a Feature or FeaturCollection from a file.
var t = require('turf')
geojsonFile = '/path/to/file/example.geojson'
t.load(geoJsonFile, function(trees, err){
if(err) throw err
console.log(trees)
})save
Saves out a feature or feature collection. 'geojson' is currently supported.
var path = './testOut/poly.geojson'
var poly = t.polygon([[[0,0], [1,0], [1,1],[0,1]]])
var type = 'geojson'
t.save(path, poly, type, function(err, res){
if(err) throw err
console.log(res) // 1
done()
})point
Creates a geojson point Feature based on an x and a y coordinate. Properties can be added optionally.
var t = require('turf')
var point1 = t.point(-75.343, 39.984)
var point2 = t.point(-75.343, 39.984, {name: 'point 1', population: 5000})
console.log(point1)
console.log(point2)linestring
Creates a geojson linestring Feature based on a coordinate array. Properties can be added optionally.
var t = require('turf')
var linestring1 = t.linestring([[102.0, -10.0], [103.0, 1.0], [104.0, 0.0], [130.0, 4.0]])
var linestring2 = t.linestring([[102.0, -10.0], [103.0, 1.0], [104.0, 0.0], [130.0, 4.0]],
{name: 'line 1', distance: 145})
console.log(linestring1)
console.log(linestring2)polygon
Creates a geojson polygon Feature based on a coordinate array. Properties can be added optionally.
var t = require('turf')
var polygon1 = t.point([[[20.0,0.0],[101.0,0.0],[101.0,1.0],[100.0,1.0],[100.0,0.0]]])
var polygon2 = t.point([[[20.0,0.0],[101.0,0.0],[101.0,1.0],[100.0,1.0],[100.0,0.0]]],
{name: 'line 1', distance: 145})
console.log(polygon1)
console.log(polygon2)featurecollection
Creates a geojson FeatureCollection based on an array of features.
var t = require('turf')
var pt1 = t.point(-75.343, 39.984, {name: 'Location A'})
var pt2 = t.point(-75.833, 39.284, {name: 'Location B'})
var pt3 = t.point(-75.534, 39.123, {name: 'Location C'})
var fc = t.featurecollection([pt1, pt2, pt3])
console.log(fc)extent
Calculates the extent of all features and returns a bounding box.
var t = require('turf')
t.load('path/to/file/example.geojson', function(err, features){
if(err) throw err
t.extent(features, function(extent){
console.log(extent) // [minX, minY, maxX, maxY]
})
})square
Calculates the minimum square bounding box for another bounding box.
var t = require('turf')
var bbox = [0,0,5,10]
t.square(bbox, function(err, square){
if(err) throw err
console.log(square) // [-2.5, 0, 7.5, 10]
})center
Calculates the absolute center point of all features.
var t = require('turf')
t.load('path/to/file/example.geojson', function(layer, err){
if(err) throw err
t.center(layer, function(center){
console.log(center)
})
})bboxPolygon
Takes a bbox and returns the equivalent polygon feature.
var t = require('turf')
var bbox = [0,0,10,10]
t.bboxPolygon(bbox, function(err, poly){
if(err) throw err
console.log(poly)
})envelope
Takes a Feature or FeatureCollection and returns a rectangular polygon feature that encompasses all vertices.
var t = require('turf')
var pt1 = t.point(-75.343, 39.984, {name: 'Location A'})
var pt2 = t.point(-75.833, 39.284, {name: 'Location B'})
var pt3 = t.point(-75.534, 39.123, {name: 'Location C'})
var fc = t.featurecollection([pt1, pt2, pt3])
t.envelope(fc, function(err, envelopePoly){
if(err) throw err
console.log(envelopePoly)
})centroid
Calculates the centroid of a polygon Feature or FeatureCollection using the geometric mean of all vertices. This lessons the effect of small islands and artifacts when calculating the centroid of a set of polygons.
var t = require('turf')
var poly = t.polygon([[[0,0], [0,10], [10,10] , [10,0]]])
t.centroid(poly, function(err, centroid){
if(err) throw err
console.log(centroid) // a point at 5, 5
})explode
Takes a Feature or FeatureCollection and return all vertices as a collection of points.
var t = require('turf')
var poly = t.polygon([[[0,0], [0,10], [10,10] , [10,0]]])
t.explode(poly, function(err, vertices){
if(err) throw err
console.log(vertices)
})combine
Combines an array of point, linestring, or polygon features into multipoint, multilinestring, or multipolygon features.
var t = require('turf')
var pt1 = t.point(50, 1)
var pt2 = t.point(100, 101)
t.combine([pt1, pt2], function(err, combined){
if(err) throw err
console.log(combined)
})inside
Checks to see if a point is inside of a polygon. The polygon can be convex or concave.
var t = require('turf')
var poly = t.polygon([[[0,0], [50, 50], [0,100], [100,100], [100,0]]])
var pt = t.point(75, 75)
t.inside(pt, poly, function(err, isInside){
if(err) throw err
console.log(isInside) // true
})buffer
Buffers a point feature to a given radius. Lines and Polygons support coming soon. Unit selection coming soon too (degrees, miles, km).
var t = require('turf')
var pt = t.point(0, 0.5)
t.buffer(pt, 10, function(err, buffered){
if(err) throw err
console.log(buffered)
})distance
Calculates the distance between two point features in degrees, radians, miles, or kilometers. This uses the haversine formula to account for global curvature.
var t = require('turf')
var point1 = t.point(-75.343, 39.984)
var point2 = t.point(-75.534, 39.123)
var unit = 'miles' // or 'kilometers', 'degrees', 'radians'
t.distance(point1, point2, unit, function(err, distance){
if(err) throw err
console.log(distance)
})nearest
Returns the nearest point feature.
var t = require('turf')
var inPoint = t.point(-75.4, 39.4, {name: 'Location A'})
var pt1 = t.point(-75.343, 39.984, {name: 'Location B'})
var pt2 = t.point(-75.833, 39.284, {name: 'Location C'})
var pt3 = t.point(-75.534, 39.123, {name: 'Location D'})
var inFeatures = t.featurecollection([pt1, pt2, pt3])
t.nearest(inPoint, inFeatures, function(err, closestPoint){
if(err) throw err
console.log(closestPoint)
})tin
Takes a set of points and the name of a z-value property and creates a tin (Triangulated Irregular Network). These are often used for developing elevation contour maps or stepped heat visualizations.
var t = require('turf')
var z = 'elevation'
t.load('/path/to/pointsfeatures/elevationPoints.geojson', function(err, points){
t.tin(points, z, function(err, tin){
if(err) throw err
console.log(tin)
})
})grid
Takes a bounding box and a cell depth and outputs a feature collection of points in a grid.
var t = require('turf')
var depth = 15
t.grid([0,0,10,10], depth, function(err, grid){
console.log(grid) // 15x15 grid of points in a FeatureCollection
})planepoint
Takes a trianglular plane and calculates the z value for a point on the plane.
var t = require('turf')
var point = t.point(-75.3221, 39.529)
// triangle is a polygon with "a", "b", and "c" values representing
// the values of the coordinates in order.
var triangle = t.polygon(
[[[-75.1221,39.57],[-75.58,39.18],[-75.97,39.86]]],
"properties": {"a": 11, "b": 122, "c": 44}
)
t.planepoint(point, triangle, function(err, zValue){
if(err) throw err
console.log(zValue)
})midpoint
Takes two point features and returns the mid point.
var t = require('turf')
var pt1 = t.point(0,0)
var pt2 = t.point(10, 0)
t.midpoint(pt1, pt2, function(err, midpoint){
if(err) throw err
console.log(midpoint)
})quantile
Takes a set of features, a property name, and a set of percentiles and outputs a quantile array. This can be passed as a break array to the contour function.
var t = require('turf')
var propertyName = 'elevation'
var percentiles = [10,30,40,60,80,90,99]
t.load('./testIn/Points3.geojson', function(err, pts){
if(err) throw err
t.quantile(pts, propertyName, percentiles, function(err, quantiles){
if(err) throw err
console.log(quantiles) // [ 12, 25, 29, 52, 76, 99, 143 ]
})
})jenks
Takes a set of features, a property name, and the desired number of breaks and outputs an array of natural breaks. This classification can be used in the contour function or for theming.
var t = require('turf')
var propertyName = 'elevation'
var num = 10
t.load('./testIn/Points3.geojson', function(err, pts){
if(err) throw err
t.jenks(pts, 'elevation', num, function(err, breaks){
if(err) throw err
done() // [ 11, 12, 18, 25, 29, 41, 50, 55, 76, 90, 143 ]
})
})contour
Takes a FeatureCollection of points with z values and an array of value breaks and generates contour polygons. This is a great way to visualize interpolated density on a map. It is often used for elevation maps, weather maps, and isocrones. The main advantage over a heat map is that contours allow you to see definitive value boundaries, and the polygons can be used to aggregate data. For example, you could get the 5000 ft elevation contour of a mountain and the 10000 ft elevation contour, then aggregate the number of trees in each to see how elevation affects tree survival.
var t = require('turf')
var z = 'elevation'
var resolution = 15
var breaks = [.1, 22, 45, 55, 65, 85, 95, 105, 120, 180]
t.load('../path/to/points.geojson', function(err, points){
t.contour(points, z, resolution, breaks, function(err, contours){
if(err) throw err
console.log(contours)
})
})Development
Run Tests
cd test
mocha .Build
sh buildWant to Contribute?
Pull requests, feature requests, comments on issues, testing, documentation, or any other type of support is welcome and encouraged. This is a big project, and I appreciate any help I can get. Let's go build a better geospatial engine for the web! Not sure where to start? Shoot me an email at morgan.herlocker [at] gmail.com or @morganherlocker.
Credits
This library is built and maintained by @morganherlocker. If you would like to contribute, please do! :)
I have taken a "picasso" approach to building this library, borrowing from existing code when available and modifying it to meet coding styles and standards of turf. Here is a list of places I have pulled ideas and/or code from (all open source or public domain, as far as I know):
https://github.com/ironwallaby/delaunay
https://github.com/jasondavies/conrec.js
http://stackoverflow.com/a/839931/461015
http://en.wikipedia.org/wiki/Haversine_formula
http://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm
https://github.com/mbloch/mapshaper
http://en.wikipedia.org/wiki/Delaunay_triangulation
http://svn.osgeo.org/grass/grass/branches/releasebranch_6_4/vector/v.overlay/main.c
http://www.ecse.rpi.edu/Homepages/wrf/Research/Short_Notes/pnpoly.html
http://en.wikipedia.org/wiki/Even%E2%80%93odd_rule
https://github.com/substack/point-in-polygon/blob/master/index.js
