from pathlib import Path
import numpy as np
from numpy.typing import NDArray
[docs]
def load_insulator_string() -> NDArray[np.void]:
r"""
Load example dataset of insulator string lifetimes with covariates.
Examples
--------
>>> from assetlife.datasets import load_insulator_string
>>> data = load_insulator_string()
>>> print(data["time"])
[70. 30. 45. ... 8.8 7.6 53. ]
>>> print(data["pHCl"])
[0.49 0.76 0.43 ... 1.12 1.19 0.35]
Returns
-------
structured array
A numpy structured array of 3 fields :
- time (``np.float64``) : observed lifetime values
- event (``np.bool_``) : boolean flag indicated if the event has been observed or not (if False, the observed lifetimes are right censored)
- entry (``np.float64``) : left truncation values
- pHCl (``np.float64``) : quantitative covariate values (concentration of pHCl)
- pH2SO4 (``np.float64``) : quantitative covariate values (concentration of pH2SO4)
- HNO3 (``np.float64``) : quantitative covariate values (concentration of HNO3)
"""
data = np.loadtxt(
Path(Path(__file__).parents[0], "csv/insulator_string.csv"),
delimiter=",",
skiprows=1,
dtype=np.dtype([
("time", np.float64),
("event", np.float64),
("entry", np.float64),
("pHCl", np.float64),
("pH2SO4", np.float64),
("HNO3", np.float64),
]),
)
# for some reason, numpy can't cast 1.0/0.0 to np.bool_
new_dtype = np.dtype([
("time", np.float64),
("event", np.bool_),
("entry", np.float64),
("pHCl", np.float64),
("pH2SO4", np.float64),
("HNO3", np.float64),
])
return data.astype(new_dtype)
[docs]
def load_circuit_breaker() -> NDArray[np.void]:
r"""
Load example dataset of circuit breaker lifetimes.
Examples
--------
>>> from assetlife.datasets import load_circuit_breaker
>>> data = load_circuit_breaker()
>>> print(data["time"])
[34. 28. 12. ... 42. 42. 37.]
Returns
-------
structured array
A numpy structured array of 3 fields :
- time (``np.float64``) : observed lifetime values
- event (``np.bool_``) : boolean flag indicated if the event has been observed or not (if False, the observed lifetimes are right censored)
- entry (``np.float64``) : left truncation values
"""
data = np.loadtxt(
Path(Path(__file__).parents[0], "csv/circuit_breaker.csv"),
delimiter=",",
skiprows=1,
dtype=np.dtype([
("time", np.float64),
("event", np.float64),
("entry", np.float64),
]),
)
# for some reason, numpy can't cast 1.0/0.0 to np.bool_
new_dtype = np.dtype([
("time", np.float64),
("event", np.bool_),
("entry", np.float64),
])
return data.astype(new_dtype)
# def load_input_turnbull() -> NDArray[np.void]:
# """_summary_
#
# Returns:
# np.ndarray: _description_
# """
# data = np.loadtxt(
# Path(Path(__file__).parents[0], "csv/input_turnbull.csv"),
# delimiter=",",
# skiprows=1,
# )
# data["event"] = data["event"].astype(np.bool_)
# return data
