Skip to main content

mapillary.utils.filter

mapillary.utils.filter#

This module contains the filter utilies for high level filtering logic

  • Copyright: (c) 2021 Facebook
  • License: MIT LICENSE

mapillary.utils.filter.by_look_at_feature(image: dict, look_at_feature: Feature)#

Filter through the given image features and return only features with the look_at_feature

  • Parameters
* **image** (*dict*) – The feature dictionary

* **look_at_feature** (*A WGS84 GIS feature**, **TurfPy*) – Feature description

  • Returns

    Whether the given feature is looking at the look_at_features

  • Return type

    bool

mapillary.utils.filter.compass_angle(data: list, angles: tuple = (0.0, 360.0))#

Filter out images that do not lie within compass angle range

  • Parameters
* **data** (*list*) – The data to be filtered

* **angles** – The compass angle range to filter through

  • Returns

    A feature list

  • Return type

    list

mapillary.utils.filter.existed_at(data: list, existed_at: str)#

Whether the first_seen_at properly existed after a specified time period

  • Parameters
* **data** (*list*) – The feature list

* **existed_at** (*str*) – The UNIX timestamp

  • Returns

    The feature list

  • Return type

    list

mapillary.utils.filter.existed_before(data: list, existed_before: str)#

Whether the first_seen_at properly existed before a specified time period

  • Parameters
* **data** (*list*) – The feature list

* **existed_before** (*str*) – The UNIX timestamp

  • Returns

    A feature list

  • Return type

    list

mapillary.utils.filter.features_in_bounding_box(data: list, bbox: dict)#

Filter for extracting features only in a bounding box

  • Parameters
* **data** (*list*) – the features list to be checked

* **bbox** (*dict*) – Bounding box coordinates
Example::
    ```python    >>> {    ...     'west': 'BOUNDARY_FROM_WEST',    ...     'south': 'BOUNDARY_FROM_SOUTH',    ...     'east': 'BOUNDARY_FROM_EAST',    ...     'north': 'BOUNDARY_FROM_NORTH'    ... }    ```

  • Returns

    Features that only exist within the bounding box selected for the given features list provided in the BBox functon

  • Return type

    list

mapillary.utils.filter.filter_values(data: list, values: list, property: str = 'value')#

Filter the features based on the existence of a specified value in one of the property.

TODO: Need documentation that lists the ‘values’, specifically, it refers to ‘value’ TODO: under ‘Detection’, and ‘Map feature’, related to issue #65

  • Parameters
* **data** (*dict*) – The data to be filtered

* **values** (*list*) – A list of values to filter by

* **property** (*str*) – The specific parameter to look into

  • Returns

    A feature list

  • Return type

    dict

mapillary.utils.filter.haversine_dist(data: dict, radius: float, coords: list, unit: str = 'm')#

Returns features that are only in the radius specified using the Haversine distance, from the haversine package

  • Parameters
* **data** (*dict*) – The data to be filtered

* **radius** (*float*) – Radius for coordinates to fall into

* **coords** (*list*) – The input coordinates (long, lat)

* **unit** (*str*) – Either ‘ft’, ‘km’, ‘m’, ‘mi’, ‘nmi’, see here [https://pypi.org/project/haversine/](https://pypi.org/project/haversine/)

  • Returns

    A feature list

  • Return type

    list

mapillary.utils.filter.hits_by_look_at(data: list, at: dict)#

Whether the given data have any feature that look at the at coordinates

  • Parameters
* **data** (*list*) – List of features with an Image entity

* **at** (*dict*) – The lng and lat coordinates
Example:
```>>> {...     'lng': 'longitude',...     'lat': 'latitude'... }```

  • Returns

    Filtered results of features only looking at at

  • Return type

    list

mapillary.utils.filter.image_type(data: list, image_type: str)#

The parameter might be ‘all’ (both is_pano == true and false), ‘pano’ (is_pano == true only), or ‘flat’ (is_pano == false only)

  • Parameters
* **data** (*list*) – The data to be filtered

* **image_type** (*str*) – Either ‘pano’ (True), ‘flat’ (False), or ‘all’ (None)

  • Returns

    A feature list

  • Return type

    list

mapillary.utils.filter.in_shape(data: list, boundary)#

Whether the given feature list lies within the shape

  • Parameters
* **data** (*list*) – A feature list to be filtered

* **boundary** – Shapely helper for determining existence of point within a boundary

  • Returns

    A feature list

  • Return type

    list

mapillary.utils.filter.is_looking_at(image_feature: Feature, look_at_feature: Feature)#

Return whether the image_feature is looking at the look_at_feature

  • Parameters
* **image_feature** (*dict*) – The feature set of the image

* **look_at_feature** (*dict*) – The feature that is being looked at

  • Returns

    Whether the diff is greater than 310, or less than 50

  • Return type

    bool

mapillary.utils.filter.max_captured_at(data: list, max_timestamp: str)#

Selects only the feature items that are less than the max_timestamp

  • Parameters
* **data** (*list*) – The feature list

* **max_timestamp** (*str*) – The UNIX timestamp as the max time

  • Returns

    Filtered feature list

  • Return type

    list

Usage:

>>> max_captured_at([{'type': 'Feature', 'geometry':... {'type': 'Point', 'coordinates': [30.98594605922699, 30.003757307208872]}, 'properties':... { ... }, ...}], '2020-05-23')

mapillary.utils.filter.min_captured_at(data: list, min_timestamp: str)#

Selects only the feature items that are less than the min_timestamp

  • Parameters
* **data** (*list*) – The feature list

* **min_timestamp** (*str*) – The UNIX timestamp as the max time

  • Returns

    Filtered feature list

  • Return type

    list

Usage:

>>> max_captured_at([{'type': 'Feature', 'geometry':... {'type': 'Point', 'coordinates': [30.98594605922699, 30.003757307208872]}, 'properties':... { ... }, ...}], '2020-05-23')

mapillary.utils.filter.organization_id(data: list, organization_id: int)#

Select only features that contain the specific organization_id

  • Parameters
* **data** (*dict*) – The data to be filtered

* **organization_id** (*int*) – The organiztion id to filter with

  • Returns

    A feature list

  • Return type

    dict

mapillary.utils.filter.pipeline(data: dict, components: list)#

A pipeline component that helps with making filtering easier. It provides access to different filtering mechanism by simplying letting users pass in what filter they want to apply, and the arguments for that filter

  • Parameters
* **data** (*dict*) – The GeoJSON to be filtered

* **components** (*list*) – The list of filters to apply

  • Returns

    The filtered feature list

  • Return type

    list

Usage:

>>> # assume variables 'data', 'kwargs'>>> pipeline(...     data=data,...     components=[...         {"filter": "image_type", "tile": kwargs["image_type"]}...         if "image_type" in kwargs...         else {},...         {"filter": "organization_id", "organization_id": kwargs["organization_id"]}...         if "organization_id" in kwargs...         else {},...         {...             "filter": "haversine_dist",...             "radius": kwargs["radius"],...             "coords": [longitude, latitude],...         }...         if "radius" in kwargs...         else 1000...     ]... )

mapillary.utils.filter.pipeline_component(func, data: list, exception_message: str, args: list)#

A pipeline component which is respnonsible for sending functional arguments over to the selected target function - throwing a warning in case of an exception

  • Parameters
* **func** (*function*) – The filter to apply

* **data** (*list*) – The list of features to filter

* **exception_message** (*str*) – The exception message to print

* **args** (*list*) – Arguments

  • Returns

    The filtered feature list

  • Return type

    list

Usage:

>>> # internally used in mapillary.utils.pipeline

mapillary.utils.filter.sequence_id(data: list, ids: list)#

Filter out images that do not have the sequence_id in the list of ids

  • Parameters
* **data** (*list*) – The data to be filtered

* **ids** (*list*) – The sequence id(s) to filter through

  • Returns

    A feature list

  • Return type

    list