Source code for libuplift.datasets.Information
"""The marketing campaign dataset from the CRAN Information package by
Kim Larsen.
See: https://cran.r-project.org/web/packages/Information/index.html
for details.
"""
import numpy as np
from .base import _fetch_remote_csv
from .base import RemoteFileMetadata
ARCHIVE_T = RemoteFileMetadata(
filename="Information_train.csv.gz",
url=('https://github.com/jszymon/uplift_sklearn_data/'
'releases/download/Information/Information_train.csv.gz'),
checksum=('7632536786357871de3f2438dbbe6c70'
'17829b2d244b2878dba9fb11af0a449b'))
ARCHIVE_V = RemoteFileMetadata(
filename="Information_valid.csv.gz",
url=('https://github.com/jszymon/uplift_sklearn_data/'
'releases/download/Information/Information_valid.csv.gz'),
checksum=('96b1773bc9fea564de03d412b936b32a'
'b5d3b1a9f7338dbb59aa6f918806fe66'))
[docs]
def fetch_Information(version="train", 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 marketing campaign dataset from the CRAN Information
package by Kim Larsen.
See: https://cran.r-project.org/web/packages/Information/index.html
Two datasets are available: "train" and "validation". Use version
argument to select.
Download it if necessary.
Parameters
----------
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_conversion, target_visit, target_exposure) : tuple if
``return_X_y`` is True
"""
# dictionaries
treatment_values = ['control', 'treated']
categ_values = dict()
# attribute descriptions
treatment_descr = [("treatment", np.int32, "TREATMENT")]
target_descr = [("target", np.int32, "PURCHASE"),
]
feature_descr = [('AGE', float),
('AGRGT_BAL_ALL_XCLD_MRTG', float),
('AUTO_2_OPEN_DATE_YRS', float),
('AUTO_HI_CRDT_2_ACTUAL', float),
('AVG_BAL_ALL_FNC_REV_ACTS', float),
('AVG_BAL_ALL_PRM_BC_ACTS', float),
('D_DEPTCARD', float),
('D_NA_AVG_BAL_ALL_FNC_REV_ACTS', float),
('D_NA_M_SNCOLDST_BNKINSTL_ACTOPN', float),
('D_NA_M_SNC_MST_RCNT_ACT_OPN', float),
('D_NA_M_SNC_MST_RCNT_MRTG_DEAL', float),
('D_NA_M_SNC_OLDST_MRTG_ACT_OPN', float),
('D_NA_RATIO_PRSNL_FNC_BAL2HICRDT', float),
('D_REGION_A', float),
('D_REGION_B', float),
('D_REGION_C', float),
('FNC_CARD_OPEN_DATE_YRS', float),
('HI_RETAIL_CRDT_LMT', float),
('MAX_MRTG_CLOSE_DATE', float),
('MRTG_1_MONTHLY_PAYMENT', float),
('MRTG_2_CURRENT_BAL', float),
('M_SNCOLDST_BNKINSTL_ACTOPN', float),
('M_SNCOLDST_OIL_NTN_TRD_OPN', float),
('M_SNC_MSTRCNT_MRTG_ACT_UPD', float),
('M_SNC_MSTREC_INSTL_TRD_OPN', float),
('M_SNC_MST_RCNT_60_DAY_RTNG', float),
('M_SNC_MST_RCNT_ACT_OPN', float),
('M_SNC_MST_RCNT_MRTG_DEAL', float),
('M_SNC_OLDST_MRTG_ACT_OPN', float),
('M_SNC_OLDST_RETAIL_ACT_OPN', float),
('N30D_ORWRS_RTNG_MRTG_ACTS', float),
('N_120D_RATINGS', float),
('N_30D_AND_60D_RATINGS', float),
('N_30D_RATINGS', float),
('N_ACTS_90D_PLS_LTE_IN_6M', float),
('N_ACTS_WITH_MXD_3_IN_24M', float),
('N_ACTS_WITH_MXD_4_IN_24M', float),
('N_BANK_INSTLACTS', float),
('N_BC_ACTS_OPN_IN_12M', float),
('N_BC_ACTS_OPN_IN_24M', float),
('N_DEROG_PUB_RECS', float),
('N_DISPUTED_ACTS', float),
('N_FNC_ACTS_OPN_IN_12M', float),
('N_FNC_ACTS_VRFY_IN_12M', float),
('N_FNC_INSTLACTS', float),
('N_INQUIRIES', float),
('N_OF_MRTG_ACTS_DLINQ_24M', float),
('N_OF_SATISFY_FNC_REV_ACTS', float),
('N_OPEN_REV_ACTS', float),
('N_PUB_REC_ACT_LINE_DEROGS', float),
('N_RETAIL_ACTS_OPN_IN_24M', float),
('N_SATISFY_INSTL_ACTS', float),
('N_SATISFY_OIL_NATIONL_ACTS', float),
('N_SATISFY_PRSNL_FNC_ACTS', float),
('PRCNT_OF_ACTS_NEVER_DLQNT', float),
('PREM_BANKCARD_CRED_LMT', float),
('RATIO_BAL_TO_HI_CRDT', float),
('RATIO_PRSNL_FNC_BAL2HICRDT', float),
('RATIO_RETAIL_BAL2HI_CRDT', float),
('STUDENT_HI_CRED_RANGE', float),
('STUDENT_OPEN_DATE_YRS', float),
('TOT_BAL_ALL_DPT_STORE_ACTS', float),
('TOT_HI_CRDT_CRDT_LMT', float),
('TOT_INSTL_HI_CRDT_CRDT_LMT', float),
('TOT_NOW_LTE', float),
('TOT_OTHRFIN_HICRDT_CRDTLMT', float),
('UPSCALE_OPEN_DATE_YRS', float),
]
if version == "train":
arch = ARCHIVE_T
elif version == "valid":
arch = ARCHIVE_V
else:
raise ValueError("Wrong Information dataset version."
" Got '{version}', expected 'train' or 'valid'.")
ret = _fetch_remote_csv(arch, "Information",
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=69, all_num=True
)
if not return_X_y:
ret.descr = __doc__
return ret