"""The actg320 trial data from Hosmer, Lemeshow and May.
"""
import numpy as np
from .base import _fetch_remote_csv
from .base import RemoteFileMetadata
ARCHIVE = RemoteFileMetadata(
filename="actg320.csv.gz",
url=('https://github.com/jszymon/uplift_sklearn_data/'
'releases/download/actg320/actg320.csv.gz'),
checksum=('bd39c16f8c15b2dada38d6af702bbcf6'
'4ccb33f0a5a8d8a95ee001a710f444c7'))
[docs]
def fetch_actg320(data_home=None,
download_if_missing=True,
random_state=None, shuffle=False,
categ_as_strings=False, return_X_y=False,
as_frame=False):
"""Load the actg320 AIDS treatment clinical trial dataset.
Download it if necessary.
This is a randomized clinical trial dataset of various AIDS
treatments from [2]_.
The description of the original study can be found in [1]_.
The main treatment variable indicates whether treatment includes
IDV (Indinavir). The treatment_grp variable contains one of four
specific treatments given:
1 = ZDV + 3TC
2 = ZDV + 3TC + IDV
3 = d4T + 3TC
4 = d4T + 3TC + IDV
(treatments 3 and 4 were given in only 3 cases)
Treatment assignment was stratified on strat2 variable (CD4 count).
Target variables:
time/censor: time/censoring to occurrence of AIDS or death
time_d/censor_d: time/censoring to occurrence of death
**Variables**
strat2
CD4 stratum at screening 0: CD4 <= 50, 1: > 50
sex
1 = Male, 2 = Female
raceth
Race/Ethnicity:
1 = White Non-Hispanic
2 = Black Non-Hispanic
3 = Hispanic (regardless of race)
4 = Asian, Pacific Islander
5 = American Indian, Alaskan Native
6 = Other/unknown
ivdrug
IV drug use history:
1 = Never
2 = Currently
3 = Previously
hemophil
Hemophiliac
karnof
Karnofsky Performance Scale
cd4
Baseline CD4 count [Cells/milliliter]
priorzdv
Months of prior ZDV use [months]
age
Age at Enrollment [years]
Parameters
----------
include_location_vars : boolean, default=True
Should variables describing hospital locations be
included. These are categorical variables with large number of
levels. The removed variables are regl, grpl, grps
data_home : string, optional
Specify another download and cache folder for the datasets. By default
all scikit-learn data is stored in '~/scikit_learn_data' subfolders.
download_if_missing : boolean, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
random_state : int, RandomState instance or None (default)
Determines random number generation for dataset shuffling. Pass an int
for reproducible output across multiple function calls.
shuffle : bool, default=False
Whether to shuffle dataset.
categ_as_strings : bool, default=False
Whether to return categorical variables as strings.
return_X_y : boolean, default=False.
If True, returns ``(data.data, data.target)`` instead of a Bunch
object.
as_frame : boolean, default=False
If True features are returned as pandas DataFrame. If False
features are returned as object or float array. Float array
is returned if all features are floats.
Returns
-------
dataset : dict-like object with the following attributes:
dataset.data : numpy array
Each row corresponds to the features in the dataset.
dataset.target : numpy array
Each value is 1 if a purchase was made 0 otherwise.
dataset.DESCR : string
Description of the dataset.
(data, target) : tuple if ``return_X_y`` is True
References
----------
.. [1] S.M. Hammer, et al., "A Controlled Trial of Two Nucleoside
Analogues plus Indinavir in Persons with Human Immunodeficiency
Virus Infection and CD4 Cell Counts of 200 per Cubic Millimeter
or Less", New England Journal of Medicine, 337(11), 725--733,
1997 (https://www.nejm.org/doi/10.1056/NEJM199709113371101)
.. [2] Hosmer, D.W. and Lemeshow, S. and May, S., Applied
Survival Analysis: Regression Modeling of Time to Event Data:
Second Edition, John Wiley and Sons Inc., New York, NY, 2008.
"""
# dictionaries
karnof_values = ["70", "80", "90", "100"]
# attribute descriptions
treatment_descr = [("treatment", np.int32, "tx"),
("treatment_grp", np.int32, "txgrp"),
]
target_descr = [("target_time", float, "time"),
("target_censor", np.int32, "censor"),
("target_time_d", float, "time_d"),
("target_censor_d", np.int32, "censor_d"),
]
feature_descr = [("strat2", np.int32),
("sex", ["1","2"]),
("raceth", ["1","2","3","4","5"]),
("ivdrug", ["1","2","3"]),
("hemophil", np.int32),
("karnof", karnof_values),
("cd4", float),
("priorzdv", float),
("age", float),
]
ret = _fetch_remote_csv(ARCHIVE, "actg320",
feature_attrs=feature_descr,
treatment_attrs=treatment_descr,
target_attrs=target_descr,
categ_as_strings=categ_as_strings,
return_X_y=return_X_y, as_frame=as_frame,
download_if_missing=download_if_missing,
random_state=random_state, shuffle=shuffle,
total_attrs=15
)
if not return_X_y:
ret.descr = __doc__
return ret