prophazCheck: Check proportional hazards assumption and log-log survival...
In buddha2490/BERG: A set of standardized functions for ACS BERG analyses

Description Usage Arguments Details Value Author(s) See Also Examples

prophazCheck quickly checks the proportional hazards assumption and returns a data frame of p-values and log-log survival plots.

1	prophazCheck(dat,start,stop,outcome,age,expo,outcome.title=NULL,expo.title=NULL)

`dat`	Data frame used in the analysis
`start`	Character vector of the variable name for your start of followup. Typically this would be "dtint92", "dtint97", etc. Variable can be formatted as a date or numeric variable. It is converted to a numeric variable within the function.
`stop`	Character vector of the variable name for your end of followup. Typically this would be "dateft", "datedd". Variable can be formatted as a date or numeric variable, but is converted to numeric within the function to calculate continuous followup time.
`outcome`	Character vector indicating the variable name for your censor variable. Variable should be coded as a numeric variable, with 0=noncase, and 1=control
`age`	Character vector indicating your baseline age variable. Cox models used the strata() function to stratify on single year of baseline age. Typically this would correspond to variables "AGE_INT", "AGE92M", etc.
`expo`	Character vector indicating your main exposure variable for which you want to assess proportional hazards.
`outcome.title`	Provide a character vector describing your censor variable. This will be used in the plots title. The function defaults to the "outcome" argument, but you can call it anything you want.
`expo.title`	Provide a character vector describing your exposure variable. This will be used in the plots title. The function defaults to the "expo" argument, but you can call it anything you want.

Proportional hazards are checked using the cox_zph() function in the survival package by using the Schoenfeld residuals against transformed time. Log-log survival plots are also calculated and returned.

A list containing 2 objects

`pval`	Data frame with three variables: exposure variable, outcome variable, and p-value for proportional hazards assumption
`plots`	log-log survival plots

Brian Carter

cox.zph, ~~~

foo <- prophazCheck(dat=example_data,
             start="dtint92",
             stop="dateft",
             outcome="myeloid",
             age="age92m",
             expo="bmicat92",
             outcome.title="All myeloid leukemias",
             expo.title="baseline BMI")
foo$plots  # draws plots

# You can save plots using the png() function

png(filename="My LogLog Plots.png", width=5,height=5,units="in",res=400)
foo$plots
dev.off()