CHANGES IN epiR VERSION 2.0.30
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.2by2 and epi.tests data entry formats simplified. Both functions now accept tables generated from tidyverse.
[Suggested by Mark Stevenson]
o Documentation for all functions reviewed. Consistency in naming of objects improved.
[Suggested by Mark Stevenson]
CHANGES IN epiR VERSION 2.0.29
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.2by2 returns NAs for point estimates of incidence rate ratio when number of events equals zero.
[Suggested by Mark Stevenson]
CHANGES IN epiR VERSION 2.0.28
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.directadj now returns zero for the directly adjusted incidence rate estimate when the number of observations in a strata equals zero and the time at risk in the same strata equals zero.
[Suggested by Mark Stevenson]
CHANGES IN epiR VERSION 2.0.26
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.2by2 now provides interpretive statements of the number needed to treat and the number needed to harm (with confidence intervals) following the approach described by Altman (1998).
[Suggested by Lucas Huggins]
o Documentation for epi.directadj updated.
CHANGES IN epiR VERSION 2.0.24
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.kappa now handles n by n tables.
[Suggested by Anita Tolpinrud]
o Inconsistency in calculation of test of significance of SMRs using Byar's method corrected.
[Spotted by Zhou Weilong]
CHANGES IN epiR VERSION 2.0.23
SIGNIFICANT USER-VISIBLE CHANGES
o New function epi.blcm.paras to return the number of unknown parameters to be inferred and the number of informative priors likely to be needed for an identifiable Bayesian latent class model to estimate diagnostic sensitivity and specificity in the absence of a gold standard.
[Contributed by Simon Firestone, Allison Cheung and Nagendra Singanallur]
o Inconsistency in calculation of test of significance of SMRs using Byar's method corrected.
[Spotted by Zhou Weilong]
CHANGES IN epiR VERSION 2.0.21
SIGNIFICANT USER-VISIBLE CHANGES
o Major tidy up for epi.2by2. Function includes argument 'interpret' to return interpretive statements for each of the computed measures of association. epi.2by2 object now includes massoc.summary, massoc.interp and massoc.detail objects.
[Suggested by Caitlin Pfeiffer and Petra Mullner]
o print(epi.2by2) now returns Yates corrected chi2 test if any of the cell frequencies are less than 5.
[Suggested by Simon Firestone]
CHANGES IN epiR VERSION 2.0.20
SIGNIFICANT USER-VISIBLE CHANGES
o rsu.dxtest code modified to check that appropriate values for test covariance have been used. Error returned when inappropriate values entered. The function now returns diagnostic sensitivity and specificity assuming tests are independent as well dependent. Documentation for rsu.dxtest updated with details on how to calculate argument covar.
[Suggested by Barbara Moloney]
CHANGES IN epiR VERSION 2.0.19
SIGNIFICANT USER-VISIBLE CHANGES
o rsu.spp.rs code modified.
o Documentation for rsu.spp.rs, epi.psi and rsu.sssep.rs updated.
CHANGES IN epiR VERSION 2.0.18
SIGNIFICANT USER-VISIBLE CHANGES
o Vignette on sample size calculations added.
CHANGES IN epiR VERSION 2.0.17
SIGNIFICANT USER-VISIBLE CHANGES
o epi.2by2 code modified to allow it handle the Haldane-Anscombe correction (i.e. addition of 0.5 to each cell of the 2 by 2 table when at least one of the cell frequencies is zero).
[Suggested by Mark Stevenson]
o Error in confidence interval calculation for epi.smr where method = "byar" corrected.
[Spotted by Sarah Haile]
o Vignette for surveillance functions updated.
o epi.2by2 calculates maximum likelihood confidence intervals for the odds ratio when the total number of events is less than 2E09, removing numeric overflow errors.
[Suggested by Stuart Reece]
o New function epi.psi: Proportional similarity index to compare non-parametric frequency distributions.
o epi.prcc now includes confidence intervals. Example provided in the documentation for epi.prcc improved.
o Function epi.ssdetect returns sample size estimates using the binomial and hypergeometric distribution.
[Suggested by Manuel Sanchez Vazquez]
CHANGES IN epiR VERSION 2.0.1
SIGNIFICANT USER-VISIBLE CHANGES
o New function epi.ssdxtest: Sample size to validate a diagnostic test in the absence of a gold standard.
o Function epi.sssimpleestb modified to adjust the required sample size to account for imperfect diagnostic sensitivity and specificity.
CHANGES IN epiR VERSION 2.0.0
SIGNIFICANT USER-VISIBLE CHANGES
o Major upgrade with amalgamation of epiR with Evan Sergeant's RSurveillance package.
o Function epi.insthaz now returns Kaplan-Meier survival estimates as well as instantaneous hazard.
o Argument nfractional added to each of the sample size functions.
o Inconsistency in reporting of results of epi.nomogram corrected.
[Spotted by Pietro Ravani]
CHANGES IN epiR VERSION 1.0-16
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.ssxsectn updated to calculate minimum detectable odds ratio for a cross-sectional study. Function epi.sscohortc updated to calculate minimum detectable odds ratio for a cohort study.
[Suggested by Alex Hou]
CHANGES IN epiR VERSION 1.0-15
SIGNIFICANT USER-VISIBLE CHANGES
o Function epi.insthaz now handles stratified data. Examples in function documentation updated to reflect these changes.
o Argument rho in function epi.sscc changed to rho.cc to avoid confusion with rho used to represent the intracluster correlation coefficient.
[Suggested by Xiaoqing Liu]
o Amibguity in documentation for epi.sscomps corrected.
[Spotted by Marco Barbara]
o Function epi.2by2 now includes Taylor series confidence intervals for the incidence risk and prevalence ratio.
CHANGES IN epiR VERSION 1.0-14
SIGNIFICANT USER-VISIBLE CHANGES
o Long overdue vignette included.
CHANGES IN epiR VERSION 1.0-13
SIGNIFICANT USER-VISIBLE CHANGES
o epi.2by2 calculates Cornfield confidence intervals for the odds ratio when the total number of events is less than 500. This reduces computation time when cell frequencies are large.
[Suggested by Jeff Canar]
o epi.ssdetect now uses the hypergeometric distribution to calculate sample size, removing the need to apply a finite correction factor.
[Suggested by Manuel Sanchez Vazquez]
o epi.ssninfb
Inconsistency in function documentation.
[Spotted by Winston Mason]
Fixed.
CHANGES IN epiR VERSION 1.0-12
SIGNIFICANT USER-VISIBLE CHANGES
o New function epi.smr: Computes confidence intervals and tests of significance of the standardised mortality [morbidity] ratio.
CHANGES IN epiR VERSION 1.0-11
SIGNIFICANT USER-VISIBLE CHANGES
o epi.2by2
Inconsistency in variables returned from print(epi.2by2) and summary(epi.2by2) corrected.
[Spotted by Jose G Conde Santiago]
CHANGES IN epiR VERSION 1.0-10
SIGNIFICANT USER-VISIBLE CHANGES
o epi.interaction
Function now includes option to use either product or dummy parameterisation of interaction terms. Documentation updated.
[Suggested by Francois M Carrier]
o New function epi.sscohortc: Sample size, power or minimum detectable incidence risk ratio for a cohort study using individual count data.
o New function epi.sscohortt: Sample size, power or minimum detectable incidence rate ratio for a cohort study using individual time data.
o New function epi.ssxsectn: Sample size, power or minimum detectable preva;ence ratio for a cross-sectional study.
CHANGES IN epiR VERSION 1.0-06
SIGNIFICANT USER-VISIBLE CHANGES
o Functions to calculate sample size. Major revision of the naming of functions to calculate sample size.
o New function epi.ssclus1estb: Sample size to estimate a binary outcome using one-stage cluster sampling.
CHANGES IN epiR VERSION 1.0-04
SIGNIFICANT USER-VISIBLE CHANGES
o epi.conf
Error in incidence rate confidence interval calculation.
[Spotted by Kazuki Yoshida]
Fixed.
o epi.detectsize
Inconsistency in function documentation.
[Spotted by Jamie Madden]
Fixed.
o epi.kappa
Function returns an error if the number of rows and number of columns in the data table presented for analysis does not equal two.
[Spotted by Maia Dolgopoloff]
Fixed.
BUG FIXES
o epi.2by
Anomalies in calculation of test of homogeneity in strata odds ratios and risk ratios corrected. Argument "homogeneity" removed from function. Woolf test of homogeneity reported by default for print(x, ...). Test statistics, degrees of freedom and p-values for Breslow Day and Woolf tests of homogeneity returned using summary(x, ...).
[Spotted by Antonio A Lopes]
Fixed.
CHANGES IN epiR VERSION 1.0-01
SIGNIFICANT USER-VISIBLE CHANGES
o epi.2by
Argument "homogeneity" removed from epi.2by2. Mantel-Haenszel (Woolf) test of homogeneity now reported by default for print(x, ...). Test statistics, degrees of freedom and p-values for Breslow Day and Woolf tests of homogeneity returned using summary(x, ...).
BUG FIXES
o epi.2by
Anomalies in calculation of test of homogeneity in strata odds ratios and risk ratios corrected. Argument "homogeneity" removed from function. Woolf test of homogeneity reported by default for print(x, ...). Test statistics, degrees of freedom and p-values for Breslow Day and Woolf tests of homogeneity returned using summary(x, ...).
[Spotted by Antonio A Lopes]
Fixed.
CHANGES IN epiR VERSION 0.9-98
SIGNIFICANT USER-VISIBLE CHANGES
o epi.prev
Function does not truncate true prevalence estimates to values between 0 and 1. Warning issued when apparent prevalence is less than (1 - specificity). Additional references added to docuementation.
CHANGES IN epiR VERSION 0.9-97
BUG FIXES
o epi.prev
Error in formatting of confidence intervals when method = "sterne" and method = "blaker".
[Spotted by Salome Duerr]
Fixed.
o epi.noninfb
Study power estimation when r argument for epi.noninfb was not equal to 1 returned incorrect results.
[Spotted by Aline Guttmann]
Fixed.
o epi.interaction
Error returned when model = coxph.
[Spotted by Eirik Degerud]
Fixed.
o epi.directadj
Inconsistency in documentation.
[Spotted by Jose G Conde Santiago]
Fixed.