readss: Read Analyze Data From ASCII File

In STAND: Statistical Analysis of Non-Detects

Description

Read data from fn.txt (space delimited text file) or fn.csv (comma delimited text file) and calculate all summary statistics using IH.summary . Output results to an ASCII text file fnout.csv in CSV format

Usage

readss(fn,L,bcol=NA,rto=5,pstat=NA,reverse=FALSE,p=0.95,gam=0.95,comma=FALSE)

Arguments

fn	name of input data file in double quotes without the .txt or .csv extension
L	L is specified limit for the exceedance fraction; e.g., the occupational exposure limit
bcol	Column that contains the BY variable–see details. Default NA
rto	Round values to rto. Default = 5
pstat	Select a subset of statistics calculated by IH.summary.Dafault All
reverse	If reverse is TRUE reverse rows and columns in output file. Default=FALSE
p	probability for Xp the 100pth percentile. Default is 0.95
gam	one-sided confidence level γ. Default is 0.95
comma	if TRUE,the input file is in csv format with column names. Default is FALSE

Details

Read data from a tab or comma delimited text file in the current folder/directory. The first column must contain measurements (observed X values). The second column is an indicator variable with 1 for a detected value and 0 for a non-detect. Additional columns can contain factors that can be used to define a BY variable. The first record in the file must contain valid R names. Valid names may contain letters (case sensitive), numbers, period, and underscore and should start with a letter ( no spaces). This file would most likely be obtained from an Excel spread sheet using the file "Save As" option, with file Save as type:
Text(Tab delimited)(*.txt) or CSV(Comma delimited)(*.csv).

Value

Returns invisible data.frame from file fn.txt

Column 1	value of measurement
Column 2	indicator variable ; 1 for detect 0 for non-detect
Column 3	… additional variables

Side effects

Summary statistics calculated by IH.summary are computed for each subset of data as defined by the levels of the BY variable. A data frame with row names from IH.summary (or subset based on value of pstat) and column names defined by the values of the BY variable is output as an ASCII text file in CSV format fnout.csv in the working folder. If reverse is TRUE the rows and columns are reversed.

Note

For information about factor see R help file factor Each level of the BY variable must have at least two non-detects for this function. If this is not the case an error message is printed to the R console and the levels of the BY variable with less than 3 non-detects are printed.

Author(s)

E. L. Frome

References

see the help file for lnorm.ml, efclnp efraction.ml, percentile.ml, kmms

See Also

About-STAND for more details and a complete reference list

Examples

# to demonstrate the use of readss add a new factor grp to the cansdata
# this factor with four levels (A_1 A_2 B_1 B_2) combines strata and sample
data(cansdata)
grp <- paste(cansdata$strata,cansdata$sample,sep="_")
temp <- data.frame(cansdata,grp) # add four level factor grp to cansdata

#    the next line is NOT executable  use CUT AND PASTE
#    sink("demoread.txt") ; print(temp) ; sink()

# The preceding line writes temp to a text file demoread.txt in the current folder
# This file would normally be created by another program, e.g. Excel
#   now use readss() to read this space delimited text file and calculate
#   all of the summary statistics for each level of grp and output
#   the results to a new text file demoreadout.csv in the current folder

#     rdemo <- readss("demoread",L=0.2,bcol=5)

#  rdemo is the R data frame that was used to calculate results in demoreadout.csv
#  to see same results rounded to three places in R console use
#  round( IH.summary(rdemo,L=0.2,bcol=5), 3)

#  To select a subset of statistics from IH.summary first define the subset
#  psel<-c("Xp.obs","Xp","Xp.UCL","f","f.UCL","Rsq","m","n")
#  entering the following command will overwrite demoreadout.csv
#  with rows and columns reversed and the subset of statistics as columns
#  and the results will be rounded to 4 places
#  rdemo <- readss("demoread",L=0.2,bcol=5,rto=4,pstat=psel,rev=TRUE)
#
#  to see same results rounded to three places in R console use
#  t(round( IH.summary(rdemo,L=0.2,bcol=5)[psel,], 3))

Example output

Loading required package: survival

STAND documentation built on May 2, 2019, 3:39 p.m.