Skip to content

Latest commit

 

History

History
118 lines (81 loc) · 8.45 KB

sink.md

File metadata and controls

118 lines (81 loc) · 8.45 KB

Sink

from tdw.proc_gen.arrangements.sink import Sink

A sink can have objects on it and inside it.

  • The sink model is chosen randomly; see Sink.MODEL_CATEGORIES["sink"].
  • The sink is placed next to a wall.
    • The sink's position is automatically adjusted to set it flush to the wall.
    • The sink is automatically rotated so that it faces away from the wall.
  • The sink will have a rectangular arrangement of objects on the counter top.
    • The objects are chosen randomly; see Sink.ON_TOP_OF["sink"].
    • The objects are positioned in a rectangular grid on the sink counter top with random rotations and positional perturbations; see Sink.COUNTER_TOP_CELL_SIZE, Sink.COUNTER_TOP_CELL_DENSITY, Sink.COUNTER_TOP_WIDTH_SCALE, and Sink.COUNTER_TOP_DEPTH_SCALE.
  • The interior of the sink may be empty; see empty in the constructor.
    • If the interior is not empty, the sink will have a rectangular arrangement of objects inside its cabinet.
      • The objects are chosen randomly; see Sink.ENCLOSED_BY["sink"].
      • The objects are positioned in a rectangular grid in the sink cabinet with random rotations and positional perturbations; see Sink.CABINET_CELL_SIZE, Sink.CABINET_CELL_DENSITY, Sink.CABINET_WIDTH_SCALE, and Sink.CABINET_DEPTH_SCALE.
  • There may be objects in the sink basin; see Sink.IN_BASIN_PROBABILITY.
    • The objects are chosen randomly; see Sink.INSIDE_OF["sink"].
    • The objects are positioned in a rectangular arrangement in the sink basin; see: Sink.BASIN_CELL_SIZE, Sink.BASIN_CELL_DENSITY, Sink.BASIN_WIDTH_SCALE, and Sink.BASIN_DEPTH_SCALE.
  • All sinks have doors that can open.
  • The root object of the sink is kinematic and the door sub-objects are non-kinematic.

Fields

  • send_commands If True, send commands when self.get_commands() is called. If False, self.get_commands() will return an empty list.

  • root_object_id The ID of the root object.

  • object_ids A list of all of the object IDs in this arrangement.

  • root_object_id The ID of the root object.

  • object_ids A list of all of the object IDs in this arrangement.

  • object_ids A list of all of the object IDs in this arrangement.

  • send_commands If True, send commands when self.get_commands() is called. If False, self.get_commands() will return an empty list.

  • root_object_id The ID of the root object.

  • object_ids A list of all of the object IDs in this arrangement.

  • root_object_id The ID of the root object.

  • object_ids A list of all of the object IDs in this arrangement.

  • object_ids A list of all of the object IDs in this arrangement.

Class Variables

Variable Type Description Value
BASIN_CELL_DENSITY float The probability from 0 to 1 of a "cell" in the sink basin rectangular arrangement being empty. Lower value = a higher density of small objects. 0.4
BASIN_CELL_SIZE float The size of each cell in the sink basin rectangular arrangement. This controls the minimum size of objects and the density of the arrangement. 0.05
BASIN_DEPTH_SCALE float When adding objects, the depth of the counter top is assumed to be actual_depth * DEPTH_SCALE. This prevents objects from being too close to the edges of the sink basin. 0.8
BASIN_WIDTH_SCALE float When adding objects, the width of the sink basin is assumed to be actual_width * WIDTH_SCALE. This prevents objects from being too close to the edges of the sink basin. 0.8
CABINET_CELL_DENSITY float The probability from 0 to 1 of a "cell" in the cabinet rectangular arrangement being empty. Lower value = a higher density of small objects. 0.1
CABINET_CELL_SIZE float The size of each cell in the cabinet rectangular arrangement. This controls the minimum size of objects and the density of the arrangement. 0.04
CABINET_DEPTH_SCALE float When adding objects, the depth of the cabinet is assumed to be actual_width * CABINET_DEPTH_SCALE. This prevents objects from being too close to the edges of the cabinet. 0.7
CABINET_WIDTH_SCALE float When adding objects, the width of the cabinet is assumed to be actual_width * CABINET_WIDTH_SCALE. This prevents objects from being too close to the edges of the cabinet. 0.7
COUNTER_TOP_CELL_DENSITY float The probability from 0 to 1 of a "cell" in the counter top rectangular arrangement being empty. Lower value = a higher density of small objects. 0.4
COUNTER_TOP_CELL_SIZE float The size of each cell in the counter top rectangular arrangement. This controls the minimum size of objects and the density of the arrangement. 0.05
COUNTER_TOP_DEPTH_SCALE float When adding objects, the depth of the counter top is assumed to be actual_depth * DEPTH_SCALE. This prevents objects from being too close to the edges of the counter top. 0.8
COUNTER_TOP_WIDTH_SCALE float When adding objects, the width of the counter top is assumed to be actual_width * WIDTH_SCALE. This prevents objects from being too close to the edges of the counter top. 0.8
DEFAULT_CELL_SIZE float The default span used for arranging objects next to each other. 0.6096
ENCLOSED_BY Dict[str, List[str]] A dictionary of categories that can be enclosed by other categories. Key = A category. Value = A list of categories of models that can enclosed by the key category. loads(Path(resource_filename(__name__, "data/enclosed_by.json")).read_text())
INSIDE_OF Dict[str, List[str]] A dictionary of categories that can be inside of other categories. Key = A category. Value = A list of categories of models that can inside of the key category. loads(Path(resource_filename(__name__, "data/inside_of.json")).read_text())
IN_BASIN_PROBABILITY float The probability (0 to 1) of there being objects in the sink. 0.7
MODEL_CATEGORIES Dict[str, List[str]] A dictionary of all of the models that may be used for procedural generation. Key = The category. Value = A list of model names. Note that this category overlaps with, but is not the same as, model_record.wcategory; see: Arrangement.get_categories_and_wcategories(). loads(Path(resource_filename(__name__, "data/models.json")).read_text())
ON_TOP_OF Dict[str, List[str]] A dictionary of categories that can be on top of other categories. Key = A category. Value = A list of categories of models that can be on top of the key category. loads(Path(resource_filename(__name__, "data/on_top_of.json")).read_text())

Functions

__init__

__init__

Sink(cabinetry, wall, corner, distance, region)

Sink(cabinetry, wall, corner, distance, region, model=None, wall_length=None, rng=None)

Parameter Type Default Description
cabinetry Cabinetry The Cabinetry set.
wall CardinalDirection The wall as a CardinalDirection that the root object is next to.
corner OrdinalDirection The origin OrdinalDirection of this wall. This is used to derive the direction.
distance float The distance in meters from the corner along the derived direction.
region InteriorRegion The InteriorRegion that the object is in.
model Union[str, ModelRecord] None Either the name of the model (in which case the model must be in models_core.json), or a ModelRecord, or None. If None, a model that fits along the wall at distance is randomly selected.
wall_length float None The total length of the lateral arrangement. If None, defaults to the length of the wall.
rng Union[int, np.random.RandomState] None Either a random seed or an numpy.random.RandomState object. If None, a new random number generator is created.

get_categories_and_wcategories

Arrangement.get_categories_and_wcategories()

(Static)

Returns: A dictionary of the categories of every model that can be used by Arrangement and their corresponding wcategory and wnid. Key = The model name. Value = A dictionary with the following keys: "category" (the ProcGenObjects category), "wcategory" (the value of record.wcategory), and "wnid" (the value of record.wnid).

get_commands

self.get_commands()

Returns: A list of commands that will generate the arrangement.

get_length

self.get_length()

Returns: The lateral extent of the object.