Occurrence Filters

Introduction

Occurrence filters are methods that take an occurrence point as input and return a binary value indicating if that point passes that particular filter. These filter functions are exposed via ‘get’ methods that allow you to configure a particular filter to fit your needs and then provide a reusable function that can be used to process entire sets of points. Filters can be chained together to “clean” occurrence data before processing.

Bounding Box Filter

A bounding box filter simply checks to see if the X,Y coordinates of a point fall within the specified bounding box.

See: get_bounding_box_filter

>>> point_1 = Point('Test species', -10, 10)
>>> point_2 = Point('Test species', 20, 20)
>>> bbox_filter = get_bounding_box_filter(0, 0, 30, 30)
>>> print(bbox_filter(point_1))
[]
>>> print(bbox_filter(point_2))
[Point('Test species', 20, 20)]

---

Disjoint Geometries Filter

A disjoint geometries filter will only pass points that fall outside of a set of geometries. This is useful if you want to do something like remove species occurrence points that are from specimens from living research collections. For example, you would create a set of bufferred point geometries around each garden locations and use those to configure the filter.

See get_disjoint_geometries_filter

>>> # Filter points not within bounding box (0, 0, 40, 40)
>>> test_geometries = ['POLYGON ((0 0, 40 0, 40 40, 0 40, 0 0))']
>>> point_1 = Point('Test species', -10, 10)
>>> point_2 = Point('Test species', 20, 20)
>>> disjoint_filter = get_disjoint_geometries_filter(test_geometries)
>>> print(disjoint_filter(point_1))
[Point('Test species', -10, 10)]
>>> print(disjoint_filter(point_2))
[]

---

Intersect Geometries Filter

The intersect geometries filter is the inverse of the disjoint geometries filter. It only passes points found within the specified test geometries.

See get_intersect_geometries_filter

>>> # Filter points within bounding box (0, 0, 40, 40)
>>> test_geometries = ['POLYGON ((0 0, 40 0, 40 40, 0 40, 0 0))']
>>> point_1 = Point('Test species', -10, 10)
>>> point_2 = Point('Test species', 20, 20)
>>> intersect_filter = get_intersect_geometries_filter(test_geometries)
>>> print(intersect_filter(point_1))
[]
>>> print(intersect_filter(point_2))
[Point('Test species', 20, 20)]

---

Unique Localities Filter

The unique localities filter is used to filter out points with identical localities. This is used to ensure that models are only produced for species with a minimum number of unique localites.

See get_unique_localities_filter

>>> point_1 = Point('Test species', -10, 10)
>>> unique_loc_filter = get_unique_localities_filter()
>>> # Should pass the first time
>>> print(unique_loc_filter(point_1))
[Point('Test species', -10, 10)]
>>> # Try a second time, should fail
>>> print(unique_loc_filter(point_1))
[]