pROC 1.5 released
pROC's steady progression goes on with version 1.5. It is avalable for R only. S+ users will need to wait for the upcoming 1.6 release which will introduce power / sample size computations.
This version introduces four new notable features:
- Variance and covariance
- Univariate Log-Concave Density Estimation smoothing
- Improvements to the plotting function
- New return values in coords
Variance and covariance
It is now possible to compute the variance of a ROC curve, and the covariance of two paired ROC curves.
library(pROC) data(aSAH) rocobj <- roc(aSAH$outcome, aSAH$s100b) var(roc1) var(roc2) cov(roc1, roc2)
Two methods are available: bootstrap, DeLong1. The bootstrap is the most versatile method. DeLong is faster but works for full AUC only. For more details, see ?var.roc
and ?cov.roc
.
Univariate Log-Concave Density Estimation smoothing
Until now, three methods were available to smooth a ROC curve: binormal
, density
and fitdistr
(to fit a distribution with MASS). Now, two new methods are available: logcondens
and logcondens.smooth
. They are based on Duembgen and Rufibach (2011)2. You first need to install the logcondens package:
install.packages("logcondens")
It doesn't need to be loaded.
plot(rocobj) rs <- smooth(rocobj, method="binormal") plot(rs, add=TRUE, col="green") rs2 <- smooth(rocobj, method="density") plot(rs2, add=TRUE, col="blue") rs3 <- smooth(rocobj, method="fitdistr", density="lognormal") plot(rs3, add=TRUE, col="magenta") rs4 <- smooth(rocobj, method="logcondens") plot(rs4, add=TRUE, col="brown") rs5 <- smooth(rocobj, method="logcondens.smooth") plot(rs5, add=TRUE, col="orange") legend("bottomright", legend=c("Empirical", "Binormal", "Density", "Log-normal", "Log-concave density", "Smoothed log-concave density"), col=c("black", "green", "blue", "magenta", "brown", "orange"), lwd=2)
Improvements to the plotting function
Several users have been bothered by the fact than in pROC (R version), the sensitivity is plotted as decreasing specificity. Most other software plot increasing 1 – specificity on the X axis. The reason is purely historical: only few statistical software can plot an axis in decreasing direction. For instance S+ cannot do it, and pROC's ROC curve are plotted as 1 – specificity there. However it makes absolutely no difference on the ROC curve itself. As it was possible, I decided plot the modern version on R rather than stick to obsolete conventions.
For those who are disturbed and prefer to stick to obsolete conventions, pROC 1.5 comes with a way to plot increasing 1 – specificity in the R version with the legacy.axes
argument.
plot(rocobj, legacy.axes=TRUE)
Note that it makes no difference to the coordinates of the plot, and if you want to add some text you still have to think in the "new" way. Consequently, the following will always be plotted to the top left corner of the curve, whatever legacy.axes
you specified:
text(1, 1, auc(rocobj), adj=c(0, 1))
New return values in coords
The ret
argument of the coords
function now accepts several new values:
- "accuracy": (sensivity + specificity) / 2
- "tn": true positives count
- "tp": true negatives count
- "fn": false negtives count (positive observations classified as negative by
predictor
) - "fp": false positives count (negative observations classified as positive by
predictor
) - "npv": negative predictive value, or tn / (tn + fn)
- "ppv": positive predictive value, or tp / (tp + fp)
In addition, sensitivity, specificity, npv and ppv can be prefixed with 1-
in order to get the opposite value. Finally two additional values are recognized:
- "npe": converted to 1-npv
- "ppe": converted to 1-ppv
Here is an example. We take the best threshold of the ROC curve rocobj
and display all the parameters of this threshold:
coords(rocobj, "best", ret=c("threshold", "specificity", "sensitivity", "accuracy", "tn", "tp", "fn", "fp", "npv", "ppv", "1-specificity", "1-sensitivity", "1-npv", "1-ppv"))
Coords also accepts a new argument: drop
to control the dimension of the return value. If drop
is FALSE
, a matrix will always be returned, even if it contains only one column. This is especially useful to make scripts more reliable.
Conclusion
Here is the full change log:
- New
cov
andvar
functions coords
accepts newret
values: "accuracy", "tn", "tp", "fn", "fp", "npv", "ppv", "1-specificity", "1-sensitivity", "1-npv", "1-ppv", "npe" and "ppe"- New
legacy.axes
argument toplot
1-specificity rather than specificity - New
axes
argument to turn off the plotting of the axis - New
logcondens
andlogcondens.smooth
(Univariate Log-Concave Density Estimation) smoothing methods - New function
has.partial.auc
to determine if an AUC is full or partial - New argument
drop
forcoords
auc
andmulticlass.auc
objects now also have secondary classnumeric
- Updated load call
- Delong's CI reversed in ROC curves with
direction=">"
- Delong's CI AUC returned values > 1 or < 0 in some rare cases
- Minor improvements in documentation
As usual, you can find the new version on ExPASy and on the CRAN. To update, type update.packages()
or install.packages("pROC")
if you want to update pROC only.
- 1. Elisabeth R. DeLong, David M. DeLong and Daniel L. Clarke-Pearson (1988) “Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach”. Biometrics 44, 837–845.
- 2. Lutz Duembgen, Kaspar Rufibach (2011) “logcondens: Computations Related to Univariate Log-Concave Density Estimation”. Journal of Statistical Software, 39, 1–28. URL: jstatsoft.org/v39/i06.
Xavier Robin
Published Monday, December 12, 2011 09:27 CET
Permalink: /blog/2011/12/12/proc-1.5-released
Tags:
pROC
Comments: 6
Comments
By Alfredo Rigalli on Tuesday, January 10, 2012 11:37 CET
I understand perfectly roc.test function, but what is the usefulness of the function cov? In addition, I was not be able to use this function with two rocobjetcs.
I have the roc objetcs: rocABC, and rocERF
when I use
cov(rocABC,rocERF) the message is
Error: is.numeric(x) || is.logical(x) is not TRUE
By Xavier on Tuesday, January 10, 2012 11:53 CET
Hi Alfredo,
The covariance is a lower level statistical information. It can be useful for other calculations (for instance it is used in the power.roc.test that will be in the next release).
Concerning the error, can you please send me the result of a traceback()
?
By Alfredo Rigalli on Saturday, January 14, 2012 14:44 CET
Xavier, I do not know exactly what is a "traceback". But I typed it at the R prompt and I obtained the following
> traceback()
3: stop(paste(ch, " is not ", if (length(r) > 1L) "all ", "TRUE",
sep = ""), call. = FALSE)
2: stopifnot(is.numeric(x) || is.logical(x), is.atomic(x))
1: cov(rocABC, rocERF)
rocABC and rocERF were obtained from the data.frame
> a
estado ABC ERF
1 poor 1 23
2 poor 2 23
3 poor 3 21
4 poor 4 14
5 poor 5 23
6 poor 1 46
7 poor 2 54
8 poor 3 34
9 poor 5 34
10 poor 5 56
11 poor 5 43
12 poor 4 65
13 poor 3 54
14 poor 2 45
15 poor 2 45
16 good 3 87
17 good 4 76
18 good 5 54
19 good 7 34
20 good 8 45
21 good 9 56
22 good 9 67
23 good 8 76
24 good 7 65
25 good 6 54
26 good 4 43
27 good 3 32
28 good 2 23
29 good 1 34
30 good 8 56
were ABC and ERF are the values of two different clinical analysis.
By Xavier on Saturday, January 14, 2012 16:57 CET
Your traceback seems to indicate that pROC is not loaded.
From the output, I see cov.roc is never called. That error message is typical of the cov function included in the stats package (base R).
Try it again after loading pROC with library(pROC)
.
By Alfredo Rigalli on Wednesday, January 18, 2012 21:33 CET
Finally pROC is running perfectly in my computer. I will send you some of my production with your development! I am really satisfied and pROC allow to finish a paper for publishing.
My problems were the consequence that when I built R tcl8.5-dev tk8.5-dev packages were not installed. So, I installed them and rebuilt R and all the functions work.
Thank you for your help.
By Xavier on Thursday, January 19, 2012 17:47 CET
I'm glad to see it solved your problem!