Movatterモバイル変換

causalPAF 1.2.1

• Added aNEWS.md file to track changes to thepackage.

causalPAF 1.2.2

• Updated the date and patch number in the description file.
• Released causalPAF package to CRAN for CRAN checks and to see ifCRAN are happy to release the package.

causalPAF 1.2.3

• After submitting 1.2.1 to Cran, I received two errors which wererectified on this version as follows. Firstly, the R examples weretaking more than 5 seconds to run in pointEstimate.R, causalPAFplot.Rand sequential_PAF.R. These require this time due to simulations andbootstraps. As a result I entered for the R examples here so that theydo not produce an error when submitting to Cran.
• Secondly, the R examples in pointEstimate.R and causalPAFplot.R hada spline as follows:“ns(apob_apoa,knots=quantile(apob_apoa,c(.25,.5,.75)),Boundary.knots=quantile(apob_apoa,c(.001,.95)))”,This was more than 106 lines long, and needed to be written on 1 line inthe R example (but did not have to be written in one line in the main Rcode and functions, since Roxygen examples added in a, new line break, ”” which was a fault with Roxygen rather than the code). As a result offeedback from CRAN, I had to make sure all occurrences of this line wereunder 101 characters in the R examples. This was solved by changing thevariable name of “apob_apoa” to “apb” in the R examples which ensuredthe line containing the spline was 100 characters or less as required toremove the CRAN submission error.

causalPAF 1.2.4

• After submitting 1.2.3 to Cran, I received feedback from Cran. Imade changes as follows based on that feedback.
• Added references into Description file.
• Replaced with for Roxygen examples, which take more than 5seconds.
• Created a smaller package dataset called ‘strokedata_smallSample’,(5,000 rows of fictional patient data) and this smaller dataset was usedin all Roxygen examples so that they take less time to run. The larger‘strokedata’ (containing 16,623 rows of fictional data is stillavailable for more accurate results).
• Updated the package number to 1.2.4 and the date to2021-10-22.
  
• When the new package data , ‘strokedata_smallSample’, was put intothe package data, it was necessary to remove an error as follows byadding in R “–resave-data” into the Build  Configure Build Tools… (whichis accessed via R studio and pressing Build and then pressing ConfigureBuild Tools ) in both of these lines: 1. Install and Restart – R CMDINSTALL additional options and 2. Build Source Package – R CMDadditional options.
• Updated doi urls in README.Rmd.
• Updated Description file to 1.2.4 and updated data to 1.2.4
• After adding in a new package dataset called‘strokedata_smallSample’ it was necessary to add Depends: R (>= 2.10)in the description file which seemed to have to do with compressing thefile.

causalPAF 1.2.4.9000

• Created a branch called ‘devel’.
• Updated the description package number for the ‘devel’ branch byadding .9000 to show it is the development branch versioni.e. 1.2.4.9000
• Updated NEWS.md for devel 1.2.4.9000 as shown here.

causalPAF 1.2.5.9000

• Updated description to 1.2.5.9000
• Made changes necessary for the new development version of CRAN Rbeing released soon as per an email from CRAN. These changes are listedbelow are relate to changes to the documentation.
• Changed Roxygen code for the following 5 files: causalPAFplot.R(parameter model_listArg definition removed dash dollar sign) ,eval_make_formula.R (parameter model_list definition removed dash dollarsign), pointEstimate.R (parameter model_listArg definition removed dashdollar sign), sim_outnode.R (parameter model_list definition removeddash dollar sign) and sim_outnode2.R (parameter model_list definitionremoved dash dollar sign). In each of these files the symbols dashdollar sign were removed since these were causing errors for linux andsome windows documentation compilation.
• Changed Roxygen code forNumBootstrap parameter definition in thefile causalPAFplot.R to read plus or minus rather than the symbols plusor minus as it may solve issue for development version of R of linux andwindows that is coming out soon that does not process documentation whenthere is a unicode minus character as per email received from CRAN.
• Removed minus sign in roxygen documentation inaddInSplinesTo_in_out.R in definitions of parameters ‘count’ and‘Subset_adjustmentSet’ to read `length( in_outDAG_SplinesRemoved[[2]] )minus count’.
• In causalPAFplot.R roxygen documentation code removed hyphen in‘maximum-likelihood’ and ‘non-Markov’ and ‘Pathway-Specific’ and‘PS-PAF’ and ‘pathway-specific’ and ‘pre-fit’.
• In checkMarkovDAG.R roxygen documentation code removed hyphen in‘non Markov’.
• In indirect_PAF_Sjolander_onesimulation.R roxygen documentation coderemoved hyphen in ‘waist hip ratio’ and ‘reweighted’.
• In overall_direct.R roxygen documentation code removed hyphen in‘waist hip ratio’
• In path_specific_onesimulation.R roxygen documentation code removedhyphen in ‘waist hip ratio’ and ‘reweighted’
• In pointestimate.R roxygen documentation code removed hyphen in‘maximum likelihood’ and ‘non Markov’ and ‘pre fit’.
• In sequential_PAF.R roxygen documentation code removed hyphen in‘Monte Carlo’ and ‘k minus 1’ and ‘non Markov’.
• In strokedata.R roxygen documentation code removed hyphen in‘preexisting’ and ‘Preclinical’ and ‘no education’ and ‘stage wise’ and‘vice versa’ and ‘waist to hip ratio’ and ‘Leisure Physical’ and ‘6.5yes or no’ and ‘disease yes or no’ and ‘2=1 to 8, 3=9 to 12’ and ‘2=1 to8, 3=9 to 12’ and ‘2=1 to 8, 3=9 to 12’ and ‘Waist to hip’ and ‘maximumlikelihood’. Changed citation to Revisiting sequential attributablefractions, J Ferguson, M O’Connell, M O’Donnell, Archives of PublicHealth, 2020.
• In strokedata_smallSample.R roxygen documentation code removedhyphen in ‘preexisting’ and ‘Preclinical’ and ‘no education’ and ‘stagewise’ and ‘vice versa’ and ‘waist to hip ratio’ and ‘Leisure Physical’and ‘6.5 yes or no’ and ‘disease yes or no’ and ‘2=1 to 8, 3=9 to 12’and ‘2=1 to 8, 3=9 to 12’ and ‘2=1 to 8, 3=9 to 12’ and ‘Waist to hip’and ‘maximum likelihood’. Changed citation to Revisiting sequentialattributable fractions, J Ferguson, M O’Connell, M O’Donnell, Archivesof Public Health, 2020.

causalPAF 1.2.5.9019

• In indirect_PAF_Sjolander_onesimulation.R andpath_specific_onesimulation.R, changed Roxygen wording for parameterweighting. In particular, changed the formula for case control weightingin roxygen documentation wording from latex formula format to wordformat as this formula was causing errors in the new R devel platformsin Windows and Linux. This addressed the errors as per email from ProfBrian D Ripley and Cran on 11th August 2022.
• Undone the change in the third bullet point above in 1.2.5.9000, byadding in the dollar sign as in each of the files causalPAFplot.R,eval_make_formula.R, pointEstimate.R , sim_outnode.R and sim_outnode2.R.
• Added the published paper Pathway-specific population attributablefractions (PS-PAFs) O’Connell and Ferguson (2022)https://doi.org/10.1093/ije/dyac079 in the descriptionfile. And updated the date and number version of the descriptionfile.

causalPAF 1.2.5

• In Description changed version from 1.2.5.9019 to 1.2.5.