class Building(object):
"""Represent a physical building (e.g. a domestic house).
Attributes
----------
utility : nilmtk Utility object
ambient : dict of nilmtk Ambient objects
Keys are pairs of the form
(<room name>, <number of that room type (indexed from 0>)
e.g. ('kitchen', 1) or ('bedroom', 3)
If there are sensors which cover the whole house (or if
a dataset gives a single set of ambient measurements for a single
house without describing which room those measurements are taken)
then use the key 'building'.
Values are nilmtk Ambient objects.
metadata : dict
geographic_coordinates : pair of floats, optional
(latitude, longitude)
Only specify this if the geo location of the building is known.
Otherwise leave this blank and users should fall back to using
the geo location specified for the dataset as a whole.
n_occupants : int, optional
Max number of occupants.
rooms : dict of strings, optional
Keys are room names. Use standard names for each room from
docs/standard_names/rooms.txt.
Values are the number of each type of room in this building.
For example:
room = {'kitchen': 1, 'bedroom': 3, ...}
original_name : string, optional
The original name for this building from the original dataset
periods_unoccupied : list, optional
A list of pd.Periods when this building was empty for more than a
day (e.g. holidays).
"""
def __init__(self):
self.metadata = {}
self.utility = Utility()
self.ambient = Ambient()
def crop(self, start, end):
"""Reduce all timeseries to just these dates"""
raise NotImplementedError
def to_json(self):
representation = copy.copy(self.metadata)
representation["utility"] = self.utility.to_json()
representation["ambient"] = self.ambient.to_json()
return json.dumps(representation)
def __init__(self):
self.metadata = {}
self.utility = Utility()
self.ambient = Ambient()
