SQUAREM

Squared Extrapolation Methods for Accelerating EM-Like Monotone Algorithms

v2026.1 · Mar 12, 2026 · GPL (>= 2)

Description

Algorithms for accelerating the convergence of slow, monotone sequences from smooth, contraction mappings such as the EM algorithm. It can be used to accelerate any smooth, linearly convergent scheme. A tutorial-style introduction is available in vignette("SQUAREM"). See also: Varadhan & Roland (2008) <doi:10.18637/jss.v092.i07>.

Downloads

154.7K

Last 30 days

275th

378.5K

Last 90 days

1.6M

Last year

Trend: +38.8% (30d vs prior 30d)

CRAN Check Status

14 OK

Show all 14 flavors

Flavor	Status	Time
r-devel-linux-x86_64-debian-clang	OK	29.9s
r-devel-linux-x86_64-debian-gcc	OK	24.4s
r-devel-linux-x86_64-fedora-clang	OK	45.6s
r-devel-linux-x86_64-fedora-gcc	OK	55.2s
r-devel-macos-arm64	OK	15s
r-devel-windows-x86_64	OK	61s
r-oldrel-macos-arm64	OK	58s
r-oldrel-macos-x86_64	OK	117s
r-oldrel-windows-x86_64	OK	63s
r-patched-linux-x86_64	OK	27.4s
r-release-linux-x86_64	OK	26.9s
r-release-macos-arm64	OK	47s
r-release-macos-x86_64	OK	114s
r-release-windows-x86_64	OK	59s

Check History

OK 14 OK · 0 NOTE · 0 WARNING · 0 ERROR · 0 FAILURE Mar 18, 2026

NOTE 0 OK · 14 NOTE · 0 WARNING · 0 ERROR · 0 FAILURE Mar 10, 2026

NOTE r-devel-linux-x86_64-debian-clang

CRAN incoming feasibility

Maintainer: ‘Ravi Varadhan <ravi.varadhan@jhu.edu>’

No Authors@R field in DESCRIPTION.
Please add one, modifying
  Authors@R: person(given = "Ravi",
                    family = "Varadhan",
                    role = c("aut", "cre"),
                    email = "ravi.varadhan@jhu.edu")
as necessary.

NOTE r-devel-linux-x86_64-debian-gcc

CRAN incoming feasibility

Maintainer: ‘Ravi Varadhan <ravi.varadhan@jhu.edu>’

No Authors@R field in DESCRIPTION.
Please add one, modifying
  Authors@R: person(given = "Ravi",
                    family = "Varadhan",
                    role = c("aut", "cre"),
                    email = "ravi.varadhan@jhu.edu")
as necessary.

NOTE r-devel-linux-x86_64-fedora-clang

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-devel-linux-x86_64-fedora-gcc

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-devel-macos-arm64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-devel-windows-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-patched-linux-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-release-linux-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-release-macos-arm64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-release-macos-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-release-windows-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-oldrel-macos-arm64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-oldrel-macos-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

NOTE r-oldrel-windows-x86_64

Rd files

checkRd: (-1) fpiter.Rd:55-58: Lost braces
    55 |     \code{tol}{ - a small, positive scalar that determines when iterations 
       |               ^
checkRd: (-1) fpiter.Rd:60-61: Lost braces
    60 |     \code{maxiter}{ - an integer denoting the
...[truncated]...
g with the corresponding log-likelihood value.  When the code{objfn} is not specified it will return the fixed-point residual instead of the objective function values. 
       |                                                                        ^

Reverse Dependencies (13)

depends

GNE

imports

BayesBrainMap INTRIGUE ashr birdie codalm lava reReg spef templateICAr tranSurv

suggests

BayesfMRI FixedPoint

Dependency Network

Version History

updated 2026.1 ← 2021.1 diff Mar 12, 2026

new 2021.1 Mar 10, 2026

updated 2021.1 ← 2020.5 diff Jan 12, 2021

updated 2020.5 ← 2020.4 diff Oct 20, 2020

updated 2020.4 ← 2020.3 diff Aug 25, 2020

updated 2020.3 ← 2020.2 diff Jun 4, 2020

updated 2020.2 ← 2020.1 diff Feb 20, 2020

updated 2020.1 ← 2017.10-1 diff Jan 12, 2020

updated 2017.10-1 ← 2016.8-2 diff Oct 7, 2017

updated 2016.8-2 ← 2016.8-1 diff Aug 17, 2016

updated 2016.8-1 ← 2014.8-1 diff Aug 11, 2016

updated 2014.8-1 ← 2012.7-1 diff Aug 28, 2014

updated 2012.7-1 ← 2010.12-1 diff Jul 25, 2012

updated 2010.12-1 ← 2010.9-1 diff Jan 2, 2011

updated 2010.9-1 ← 2010.8-2 diff Sep 25, 2010

updated 2010.8-2 ← 2010.8-1 diff Aug 8, 2010

new 2010.8-1 Aug 7, 2010