episensr 2.0.0

• Using NORTA transformation, replacing sampling from a standard uniform distribution, for correlations between sensitivity and specificity of exposure classification among cases and controls in Probabilistic Bias Analyses (PBA).
• Replacing log-logistic and logit-normal distributions by a truncated normal distribution in PBA (package truncnorm).
• Due to these breaking changes in PBA functions, legacy versions are also available but will be removed in future versions.
• Remove Or.sel from multidimBias.

episensr 1.3.0

• Adding bias-adjusting for exposure misclassification using predictive values.
• Adding start-up message.

episensr 1.2.0

• Fix CI for OR in plot.booted (@codiewood, #2)
• Replaced dot-dot notation by after_stat(density) in plot.probsens and plot.booted
• misclassification_cov replaced by misclassification.cov
• Standard output for confounders.ext, confounders.limit, and confounders.array
• Suggest tidyr instead of tidyverse

episensr 1.1.0

• Fix confidence interval values in plotting output of probsens.
• Fix bug that created integer overflow in rare cases when 2-by-2 table cells were very large.
• Update documentation.

episensr 1.0.0

• Provide default values for variables in mbias.
• Update M-bias plot by using ggdag (and dagitty).
• Add beta distribution to probabilistic bias analyses.
• Plotting functions for probabilistic bias analyses.
• Review criteria to remove negative cell counts in probabilistic bias analyses.
• Add citation and DOI.

episensr 0.9.6

• Fix formulas in probsens.conf.

episensr 0.9.5

• Fix selection-bias factor in selection which was returning a constant value.

episensr 0.9.4

• Fix bias_parms in selection to use a single selection-bias factor (was skipping it and returning NAs).

episensr 0.9.3

• Update description of parms in confounders, confounders.emm, and confounders.poly
• New function confounders.array, sensitivity analysis for unmeasured confounders based on confounding imbalance among exposed and unexposed (Schneeweiss, 2006)
• Deprecated bias parameters removed from function multidimBias. Please use bias_parms instead.
• New function confounders.evalue, computing E-value to assess bias due to unmeasured confounder (VanderWeele and Ding, 2017)
• New function multiple.bias allowing to extract 2-by-2 table from an episensr object to feed another function (multiple bias analysis)

episensr 0.9.2

• Fix bug for distributions and computations of OR/RR in probsens.conf
• Update distributions in probsens.irr.conf
• Add example using probsens.conf in vignette

episensr 0.9.1

• Fix bug when using triangular distribution in probsens.conf function for prevalence of exposure among the non-exposed (as producing NaNs) (#1).

episensr 0.9.0

• Add misclassification_cov for a misclassified covariate (confounder or effect modifier).
• As such, this (wrongly) added option into misclassification in the previous version is now removed.
• Add computation of confidence interval for odds ratio as per Chu et al. for exposure misclassification.
• Use bias_parms instead of bias in misclassification function.

episensr 0.8.0

• Fix bug when building 2-by-2 table.

• Various formatting improvements in output of confounders, confounders.emm, misclassification and selection functions.

• Standardize use of bias_parms.

• Add vignette.

• Selection bias factor now available in output of selection function.

• Add bootstrap option

episensr 0.7.2

• Fix 2-by-2 tables when variables are provided instead of a matrix.

episensr 0.7.1

• Fix R version dependency (R >= 3.2.0)

episensr 0.7.0

• Harmonization of arguments across functions.

• New distributions added to probsens series of functions: constant, logit-logistic, logit-normal, log-logistic, and log-normal.

• Probabilistic analysis of person-time data added with probsens.irr for exposure misclassification, and probsens.irr.conf for unmeasured confounder.

• Sensitivity analysis to correct for selection bias caused by M bias with mbias function, including DAG plot and print function.

• Fix CI formatting.

• NAMESPACE: add imports to stats functions to avoid new R CMD CHECK warnings