From f61d96c3129ce462f6cf10a5de596167817a60c9 Mon Sep 17 00:00:00 2001
From: Brandon Carter <b-carter@users.noreply.github.com>
Date: Mon, 19 Dec 2016 17:27:22 -0500
Subject: [PATCH] [MRG + 1] Fix reference in fetch_kddcup99 (#8071)
* fix 'percent10' parameter default in fetch_kddcup99 docstring
* Consistent default 'percent10' value in _fetch_brute_kddcup99 to be consistent
---
sklearn/datasets/kddcup99.py | 16 +++++++++-------
1 file changed, 9 insertions(+), 7 deletions(-)
diff --git a/sklearn/datasets/kddcup99.py b/sklearn/datasets/kddcup99.py
index 03bf3f8d8f..c2ed39caa1 100644
--- a/sklearn/datasets/kddcup99.py
+++ b/sklearn/datasets/kddcup99.py
@@ -45,7 +45,7 @@ def fetch_kddcup99(subset=None, shuffle=False, random_state=None,
The KDD Cup '99 dataset was created by processing the tcpdump portions
of the 1998 DARPA Intrusion Detection System (IDS) Evaluation dataset,
- created by MIT Lincoln Lab [1] . The artificial data was generated using
+ created by MIT Lincoln Lab [1]. The artificial data was generated using
a closed network and hand-injected attacks to produce a large number of
different types of attack with normal activity in the background.
As the initial goal was to produce a large training set for supervised
@@ -134,7 +134,7 @@ def fetch_kddcup99(subset=None, shuffle=False, random_state=None,
shuffle : bool, default=False
Whether to shuffle dataset.
- percent10 : bool, default=False
+ percent10 : bool, default=True
Whether to load only 10 percent of the data.
download_if_missing : bool, default=True
@@ -155,9 +155,11 @@ def fetch_kddcup99(subset=None, shuffle=False, random_state=None,
Detection Evaluation Richard Lippmann, Joshua W. Haines,
David J. Fried, Jonathan Korba, Kumar Das
- .. [2] A Geometric Framework for Unsupervised Anomaly Detection: Detecting
- Intrusions in Unlabeled Data (2002) by Eleazar Eskin, Andrew Arnold,
- Michael Prerau, Leonid Portnoy, Sal Stolfo
+ .. [2] K. Yamanishi, J.-I. Takeuchi, G. Williams, and P. Milne. Online
+ unsupervised outlier detection using finite mixtures with
+ discounting learning algorithms. In Proceedings of the sixth
+ ACM SIGKDD international conference on Knowledge discovery
+ and data mining, pages 320-324. ACM Press, 2000.
"""
kddcup99 = _fetch_brute_kddcup99(shuffle=shuffle, percent10=percent10,
@@ -214,7 +216,7 @@ def fetch_kddcup99(subset=None, shuffle=False, random_state=None,
def _fetch_brute_kddcup99(subset=None, data_home=None,
download_if_missing=True, random_state=None,
- shuffle=False, percent10=False):
+ shuffle=False, percent10=True):
"""Load the kddcup99 dataset, downloading it if necessary.
@@ -242,7 +244,7 @@ def _fetch_brute_kddcup99(subset=None, data_home=None,
shuffle : bool, default=False
Whether to shuffle dataset.
- percent10 : bool, default=False
+ percent10 : bool, default=True
Whether to load only 10 percent of the data.
Returns
--
GitLab