ST_FrechetDistance

ST_FrechetDistance takes two linestring columns and returns a double column. The output column represents the discrete Fréchet distance between the two input linestrings.

The calculation pairs each point in the first linestring with a point in the second linestring as you move through both in order. The discrete Fréchet distance is the smallest possible value of the largest distance between any such paired points as you walk along both linestrings in order.

For more details on the discrete Fréchet distance calculation, see Computing Discrete Fréchet Distance.

The result will be in the same units as the input geometry data. For example, if the input geometries are in a spatial reference that uses meters, the result values will be in meters.

If the two geometry columns are in different spatial references, the function will automatically transform the second geometry into the spatial reference of the first.

If one of the input geometries has an unknown spatial reference the function will use planar distance calculations.

ST_FrechetDistance does not take m-values and z-values into account. If the input linestring has any of these values, they will be ignored in the Fréchet distance calculation.

ST_FrechetDistance can be used to measure the similarity between two linestrings. For more details, go to the core topic for similarity measures.

For more details, go to the GeoAnalytics Engine API reference for frechet_distance.

Examples

1
2
3
4
5
6
7
8
9
10
11
12
13

from geoanalytics.sql import functions as ST, Linestring

data = [
    ("LINESTRING (10 10, 20 20)", "LINESTRING (10 30, 20 40)"),
    ("LINESTRING (0 20, 10 20)", "LINESTRING (0 80, -10 20)")
]

df = spark.createDataFrame(data, ["line1_wkt", "line2_wkt"]) \
          .withColumn("line1", ST.line_from_text("line1_wkt", 4326)) \
          .withColumn("line2", ST.line_from_text("line2_wkt", 4326))

df.select(ST.frechet_distance("line1", "line2").alias("frechet_distance")).show()

1
2
3
4
5
6
+----------------+
|frechet_distance|
+----------------+
|            20.0|
|            60.0|
+----------------+

Version table

Related topics

• SQL functions
• ST_Distance
• ST_GeodesicDistance
• ST_EuclideanDistance
• ST_HausdorffDistance
• Data sources
• Coordinate systems and transformations