OutlierDC-package: Functions for detecting outlying observations for censored...
In OutlierDC: Outlier Detection using quantile regression for Censored Data
Description
Details
Note
Author(s)
References
See Also
Examples
Description
This package offers three outlier detection algorithms for censored data using quantile regression.
Details
Package: OutlierDC
Type: Package
Version: 0.3-0
Date: 2014-03-23
License: GPL version 3
LazyLoad: no
Note
We would like to thank Huxia Judy Wang and Lan Wang for permission to use their LCRQ functions.
Author(s)
Soo-Heang Eo and HyungJun Cho
Maintainer: Soo-Heang Eo <eo.sooheang@gmail.com>
References
Eo S-H, Hong S-M Hong, Cho H (2014). Identification of outlying observations with quantile regression for censored data, Submitted.
Wang HJ, Wang L (2009) Locally Weighted Censored Quantile Regression. JASA 104:1117–1128. doi: 10.1198/jasa.2009.tm08230
See Also
odc, plot, coef, show, quantreg
Examples
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
## Not run:
require(OutlierDC)
# Toy example
data(ebd)
# The data consists of 402 observations with 6 variables.
dim(ebd)
# To show the first six observations of the dataset,
head(ebd)
#scoring algorithm
fit <- odc(Surv(log(time), status) ~ meta, data = ebd)
fit
coef(fit)
plot(fit)
# Add upper bound for the selection of outleirs
fit1 <- update(fit, k_s = 4)
fit1
plot(fit1)
# residual-based algorithm
fit2 <- odc(Surv(log(time), status) ~ meta, data = ebd, method = "residual", k_r = 1.5)
fit2
plot(fit2)
# To display all of outlying observations in the fitted object
fit2@outlier.data
# boxplot algorithm
fit3 <- odc(Surv(log(time), status) ~ meta, data = ebd, method = "boxplot", k_b = 1.5)
fit3
plot(fit3, ylab = "log survival times", xlab = "metastasis lymph nodes")
## End(Not run)
Example output
Loading required package: survival
Loading required package: quantreg
Loading required package: SparseM
Attaching package: 'SparseM'
The following object is masked from 'package:base':
backsolve
Attaching package: 'quantreg'
The following object is masked from 'package:survival':
untangle.specials
Loading required package: Formula
Package OutlierDC (0.3-0) loaded.
[1] 402 6
id meta exam status time ratio
1787 55468952 0 12 1 26 0.0000000
1788 8883016 0 12 1 11 0.0000000
1789 10647194 0 12 0 134 0.0000000
1790 16033679 2 12 1 1 0.1666667
1791 19519884 0 12 0 111 0.0000000
1792 19574077 0 12 1 8 0.0000000
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd)
Algorithm: Scoring algorithm (score)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_s:
# of outliers detected: 0
Top 6 outlying scores:
times delta (Intercept) meta score Outlier
346 4.48 0 1 9 4.59
327 2.71 1 1 13 4.54
326 2.08 1 1 14 2.52
296 4.86 1 1 4 2.35
354 3.09 1 1 10 2.11
233 5.29 0 1 1 1.95
q10 q25 q50 q75 q90
(Intercept) 1.632 2.565 3.401 4.500 5.196
meta -0.022 -0.077 -0.111 -0.183 -0.191
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd)
Algorithm: Scoring algorithm (score)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_s: 4
# of outliers detected: 2
Top 6 outlying scores:
times delta (Intercept) meta score Outlier
346 4.48 0 1 9 4.59 *
327 2.71 1 1 13 4.54 *
326 2.08 1 1 14 2.52
296 4.86 1 1 4 2.35
354 3.09 1 1 10 2.11
233 5.29 0 1 1 1.95
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd, method = "residual",
k_r = 1.5)
Algorithm: Residual-based algorithm (residual)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_r: 1.5
# of outliers detected: 9
Outliers detected:
times delta (Intercept) meta residual sigma Outlier
57 4.80 0 1 2 1.63 1.6 *
80 5.04 1 1 0 1.64 1.6 *
189 5.38 0 1 0 1.98 1.6 *
191 5.20 0 1 0 1.80 1.6 *
233 5.29 0 1 1 2.00 1.6 *
296 4.86 1 1 4 1.90 1.6 *
6 of all 9 outliers were displayed.
id meta exam status time ratio
57 39165334 2 13 0 122 0.15384615
80 2022934 0 13 1 154 0.00000000
189 25678892 0 16 0 217 0.00000000
191 10521031 0 17 0 181 0.00000000
233 52223267 1 18 0 198 0.05555556
296 27085350 4 20 1 129 0.20000000
346 12269804 9 24 0 88 0.37500000
357 17822095 0 25 1 157 0.00000000
395 43506173 0 37 0 152 0.00000000
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd, method = "boxplot",
k_b = 1.5)
Algorithm: Boxplot algorithm (boxplot)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_b: 1.5
# of outliers detected: 1
Outliers detected:
times delta (Intercept) meta UB Outlier
346 4.48 0 1 9 4.32 *
1 of all 1 outliers were displayed.
OutlierDC documentation built on May 1, 2019, 11:31 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Description Details Note Author(s) References See Also Examples
Description
This package offers three outlier detection algorithms for censored data using quantile regression.
Details
| Package: | OutlierDC |
| Type: | Package |
| Version: | 0.3-0 |
| Date: | 2014-03-23 |
| License: | GPL version 3 |
| LazyLoad: | no |
Note
We would like to thank Huxia Judy Wang and Lan Wang for permission to use their LCRQ functions.
Author(s)
Soo-Heang Eo and HyungJun Cho Maintainer: Soo-Heang Eo <eo.sooheang@gmail.com>
References
Eo S-H, Hong S-M Hong, Cho H (2014). Identification of outlying observations with quantile regression for censored data, Submitted.
Wang HJ, Wang L (2009) Locally Weighted Censored Quantile Regression. JASA 104:1117–1128. doi: 10.1198/jasa.2009.tm08230
See Also
odc, plot, coef, show, quantreg
Examples
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 | ## Not run:
require(OutlierDC)
# Toy example
data(ebd)
# The data consists of 402 observations with 6 variables.
dim(ebd)
# To show the first six observations of the dataset,
head(ebd)
#scoring algorithm
fit <- odc(Surv(log(time), status) ~ meta, data = ebd)
fit
coef(fit)
plot(fit)
# Add upper bound for the selection of outleirs
fit1 <- update(fit, k_s = 4)
fit1
plot(fit1)
# residual-based algorithm
fit2 <- odc(Surv(log(time), status) ~ meta, data = ebd, method = "residual", k_r = 1.5)
fit2
plot(fit2)
# To display all of outlying observations in the fitted object
fit2@outlier.data
# boxplot algorithm
fit3 <- odc(Surv(log(time), status) ~ meta, data = ebd, method = "boxplot", k_b = 1.5)
fit3
plot(fit3, ylab = "log survival times", xlab = "metastasis lymph nodes")
## End(Not run)
|
Example output
Loading required package: survival
Loading required package: quantreg
Loading required package: SparseM
Attaching package: 'SparseM'
The following object is masked from 'package:base':
backsolve
Attaching package: 'quantreg'
The following object is masked from 'package:survival':
untangle.specials
Loading required package: Formula
Package OutlierDC (0.3-0) loaded.
[1] 402 6
id meta exam status time ratio
1787 55468952 0 12 1 26 0.0000000
1788 8883016 0 12 1 11 0.0000000
1789 10647194 0 12 0 134 0.0000000
1790 16033679 2 12 1 1 0.1666667
1791 19519884 0 12 0 111 0.0000000
1792 19574077 0 12 1 8 0.0000000
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd)
Algorithm: Scoring algorithm (score)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_s:
# of outliers detected: 0
Top 6 outlying scores:
times delta (Intercept) meta score Outlier
346 4.48 0 1 9 4.59
327 2.71 1 1 13 4.54
326 2.08 1 1 14 2.52
296 4.86 1 1 4 2.35
354 3.09 1 1 10 2.11
233 5.29 0 1 1 1.95
q10 q25 q50 q75 q90
(Intercept) 1.632 2.565 3.401 4.500 5.196
meta -0.022 -0.077 -0.111 -0.183 -0.191
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd)
Algorithm: Scoring algorithm (score)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_s: 4
# of outliers detected: 2
Top 6 outlying scores:
times delta (Intercept) meta score Outlier
346 4.48 0 1 9 4.59 *
327 2.71 1 1 13 4.54 *
326 2.08 1 1 14 2.52
296 4.86 1 1 4 2.35
354 3.09 1 1 10 2.11
233 5.29 0 1 1 1.95
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd, method = "residual",
k_r = 1.5)
Algorithm: Residual-based algorithm (residual)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_r: 1.5
# of outliers detected: 9
Outliers detected:
times delta (Intercept) meta residual sigma Outlier
57 4.80 0 1 2 1.63 1.6 *
80 5.04 1 1 0 1.64 1.6 *
189 5.38 0 1 0 1.98 1.6 *
191 5.20 0 1 0 1.80 1.6 *
233 5.29 0 1 1 2.00 1.6 *
296 4.86 1 1 4 1.90 1.6 *
6 of all 9 outliers were displayed.
id meta exam status time ratio
57 39165334 2 13 0 122 0.15384615
80 2022934 0 13 1 154 0.00000000
189 25678892 0 16 0 217 0.00000000
191 10521031 0 17 0 181 0.00000000
233 52223267 1 18 0 198 0.05555556
296 27085350 4 20 1 129 0.20000000
346 12269804 9 24 0 88 0.37500000
357 17822095 0 25 1 157 0.00000000
395 43506173 0 37 0 152 0.00000000
Please wait...
Done.
Outlier Detection for Censored Data
Call: odc(formula = Surv(log(time), status) ~ meta, data = ebd, method = "boxplot",
k_b = 1.5)
Algorithm: Boxplot algorithm (boxplot)
Model: Locally weighted censored quantile regression (Wang)
Value for cut-off k_b: 1.5
# of outliers detected: 1
Outliers detected:
times delta (Intercept) meta UB Outlier
346 4.48 0 1 9 4.32 *
1 of all 1 outliers were displayed.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.