TagCollection¶

class supervisely_lib.annotation.tag_collection.TagCollection(items=None)[source]¶

Bases: supervisely_lib.collection.key_indexed_collection.MultiKeyIndexedCollection

Collection with Tag instances. TagCollection object is immutable.

Usage example

# Create TagMetas (see class TagMeta for additional information about creating TagMetas)
meta_weather = sly.TagMeta('Weather', sly.TagValueType.ANY_STRING)

seasons = ["Winter", "Spring", "Summer", "Autumn"]
meta_season = sly.TagMeta('Season', sly.TagValueType.ONEOF_STRING, possible_values=seasons)

# Create Tags
tag_weather = sly.Tag(meta_weather, value="Sunny")
tag_season = sly.Tag(meta_season, value="Spring")
tags_arr = [tag_weather, tag_season]

# Create TagCollection from Tags
tags = sly.TagCollection(tags_arr)

# Add item to TagCollection
meta_potato = sly.TagMeta('potato', sly.TagValueType.NONE)
tag_potato =sly.Tag(meta_potato)

# Remember that TagCollection is immutable, and we need to assign new instance of TagCollection to a new variable
tags = tags.add(tag_potato)

# You can also add multiple items to collection
meta_cabbage = sly.TagMeta('cabbage', sly.TagValueType.NONE)
meta_carrot = sly.TagMeta('carrot', sly.TagValueType.NONE)
meta_turnip = sly.TagMeta('turnip', sly.TagValueType.NONE)

tag_cabbage = sly.Tag(meta_cabbage)
tag_carrot = sly.Tag(meta_carrot)
tag_turnip = sly.Tag(meta_turnip)

additional_veggies = [tag_cabbage, tag_carrot, tag_turnip]

tags = tags.add_items(additional_veggies)

# Has key, checks if given key exist in collection
tags.has_key("cabbage")
# Output: True

# Intersection, finds intersection of given list of instances with collection items
meta_dog = sly.TagMeta('dog', sly.TagValueType.NONE)
tag_dog = sly.Tag(meta_dog)

meta_cat = sly.TagMeta('cat', sly.TagValueType.NONE)
tag_cat = sly.Tag(meta_cat)

meta_turtle = sly.TagMeta('turtle', sly.TagValueType.NONE)
tag_turtle = sly.Tag(meta_turtle)


tags_animals = sly.TagCollection([tag_dog, tag_cat, tag_turtle])

tags_intersections = tags.intersection(tags_animals)
print(tags_intersections.to_json())
# Output: []

# Let's add the potato Tag from another collection and compare them again
tags_animals = tags_animals.add(tag_potato)

tags_intersections = tags.intersection(tags_animals)
print(tags_intersections.to_json())
# Output: [
#     {
#         "name":"potato"
#     }
# ]

# Difference, finds difference between collection and given list of Tags or TagCollection
meta_car = sly.TagMeta('car', sly.TagValueType.NONE)
tag_car = sly.Tag(meta_car)

meta_bicycle = sly.TagMeta('bicycle', sly.TagValueType.NONE)
tag_bicycle = sly.Tag(meta_bicycle)

tags_vehicles = sly.TagCollection([tag_car, tag_bicycle])

meta_pedestrian = sly.TagMeta('pedestrian', sly.TagValueType.NONE)
tag_pedestrian = sly.Tag(meta_pedestrian)

meta_road = sly.TagMeta('road', sly.TagValueType.NONE)
tag_road = sly.Tag(meta_road)

difference = tags_vehicles.difference([tag_pedestrian, tag_road])
print(difference.to_json())
# Output: [
#     {
#         "name":"car"
#     },
#     {
#         "name":"bicycle"
#     }
# ]

# Merge, merges collection and given list of TagMetas
tags_vehicles = sly.TagMetaCollection([tag_car, tag_bicycle])
tags_merge = sly.TagMetaCollection([tag_pedestrian, tag_road])

merged_collections = tags_vehicles.merge(tags_merge)
print(merged_collections.to_json())
# Output: [
#     {
#         "name":"pedestrian"
#     },
#     {
#         "name":"road"
#     },
#     {
#         "name":"car"
#     },
#     {
#         "name":"bicycle"
#     }
# ]

item_type¶: alias of supervisely_lib.annotation.tag.Tag

to_json() → List[dict][source]¶

Convert the TagCollection to a list of json dicts. Read more about Supervisely format.

Returns

List of dicts in json format

Return type

List[dict]

Usage example

meta_weather = sly.TagMeta('Weather', sly.TagValueType.ANY_STRING)

seasons = ["Winter", "Spring", "Summer", "Autumn"]
meta_season = sly.TagMeta('Season', sly.TagValueType.ONEOF_STRING, possible_values=seasons)

tag_weather = sly.Tag(meta_weather, value="Sunny")
tag_season = sly.Tag(meta_season, value="Spring")

tags_arr = [tag_weather, tag_season]
tags = sly.TagCollection(tags_arr)

tags_json = tags.to_json()
# Output: [
#     {
#         "name":"Weather",
#         "value":"Sunny"
#     },
#     {
#         "name":"Season",
#         "value":"Spring"
#     }
# ]

classmethod from_json(data: List[dict], tag_meta_collection: supervisely_lib.annotation.tag_meta_collection.TagMetaCollection) → supervisely_lib.annotation.tag_collection.TagCollection [source]¶

Convert a list with dicts in json format to TagCollection. Read more about Supervisely format.

Parameters

data (List[dict]) – List with dicts in json format.
tag_meta_collection (TagMetaCollection) – Input TagMetaCollection object.

Returns

TagCollection object

Return type

TagCollection

Usage example

# Initialize TagMetaCollection

meta_weather = sly.TagMeta('Weather', sly.TagValueType.ANY_STRING)

seasons = ["Winter", "Spring", "Summer", "Autumn"]
meta_season = sly.TagMeta('Season', sly.TagValueType.ONEOF_STRING, possible_values=seasons)

metas_arr = [meta_weather, meta_season]
tag_metas = sly.TagMetaCollection(metas_arr)

data = [
    {
        "name":"Weather",
        "value":"Sunny"
    },
    {
        "name":"Season",
        "value":"Spring"
    }
]

tags = sly.TagCollection.from_json(data, tag_metas)