Package: NMAoutlier
Title: Detecting Outliers in Network Meta-Analysis
Version: 0.1.17
Date: 2021-01-31
Depends: R (>= 3.0.0)
Imports: netmeta (>= 0.9-7), stats (>= 3.4.3), parallel (>= 3.4.1),
        MASS (>= 7.3-47), reshape2 (>= 1.4.3), ggplot2 (>= 3.0.0),
        gridExtra (>= 2.3)
Authors@R: c(person("Maria", "Petropoulou",
                    role = c("aut", "cre"),
                    email = "petropoulou@imbi.uni-freiburg.de",
                    comment = c(ORCID = "0000-0002-7147-3644")),
             person("Guido", "Schwarzer",
                    role = "aut",
                    comment = c(ORCID = "0000-0001-6214-9087")),
             person("Agapios", "Panos",
                    role = "aut"),
             person("Dimitris", "Mavridis",
                    role = "aut",
                    comment = c(ORCID = "0000-0003-1041-4592")))         
Author: Maria Petropoulou [aut, cre] (<https://orcid.org/0000-0002-7147-3644>),
  Guido Schwarzer [aut] (<https://orcid.org/0000-0001-6214-9087>),
  Agapios Panos [aut],
  Dimitris Mavridis [aut] (<https://orcid.org/0000-0003-1041-4592>)
Maintainer: Maria Petropoulou <petropoulou@imbi.uni-freiburg.de>
URL: https://github.com/petropouloumaria/NMAoutlier
Description: A set of functions providing several outlier (i.e., studies with extreme findings) and influential detection measures and methodologies in network meta-analysis:
               - Simple outlier and influential deletion measures provided: (a) Raw, (b) Standardized, (c) Studentized residuals; (d) Mahalanobis distance and (c) leverage.
               - Outlier and influential detection measures by considering study deletion (Shift the mean): (a) Raw (b) Standardized, (c) Studentized deleted residuals; 
                 (d) Cook distance; (e) Ratio of variance-covariance matrix; (f) weight leave one out; (g) leverage leave one out; (h) heterogeneity leave one out; 
                 (i) R heterogeneity; (k) R Q total; (l) R Q heterogeneity (m);  R Q inconsistency and (n) statistic that indicate the effect that deleting each study has on the treatment estimates.
               - Plots for outlier and influential detection simply and deletion measures (all the above measures) and Q-Q plot for network meta-analysis.
               - Forward Search algorithm in network meta-analysis. 
               - Forward plots for the monitoring statistics in each step of Forward search algorithm:
                 (a) P-scores; (b) z-values for difference of direct and indirect evidence with back-calculation method; 
                 (c) Standardized residuals; (d) heterogeneity variance estimator; (e) cook distance; (f) ratio of variances; 
                 (g) Q statistics.
               - Forward plots for summary estimates and their confidence intervals in each step of forward search algorithm.   
License: GPL (>= 2)
Encoding: UTF-8
RoxygenNote: 7.1.1
NeedsCompilation: no
Packaged: 2021-02-01 09:06:37 UTC; Maria
Repository: CRAN
Date/Publication: 2021-02-01 15:20:11 UTC
