estimate_read_csv: Read an Estimate from CSV - File.

In decisionSupport: Quantitative Support of Decision Making under Uncertainty

Read an Estimate from CSV - File.

Description

This function reads an estimate from the specified csv files. In this context, an estimate of several variables is defined by its marginal distribution types, its marginal 90%-confidence intervals [lower,upper] and, optionally, its correlations.

estimate_read_csv_old reads an estimate from CSV file(s) according to the deprecated standard. This function is for backward compatibility only.

Usage

estimate_read_csv(fileName, strip.white = TRUE, ...)

estimate_read_csv_old(fileName, strip.white = TRUE, ...)

Arguments

fileName	Name of the file containing the marginal information of the estimate that should be read.
strip.white	logical. Used only when sep has been specified, and allows the stripping of leading and trailing white space from unquoted character fields (numeric fields are always stripped). See scan for further details (including the exact meaning of ‘white space’), remembering that the columns may include the row names.
...	Further parameters to be passed to read.table.

Details

An estimate might consists of uncorrelated and correlated variables. This is reflected in the input file structure, which is described in the following.

CSV input file structures

The estimate is read from one or two csv files: the marginal csv file which is mandatory and the correlation csv file which is optional. The marginal csv file contains the definition of the distribution of all variables ignoring potential correlations. The correlation csv file only defines correlations.

The structure of the marginal distributions input file (mandatory)

File name structure: <marginal-filename>.csv

Mandatory columns:

Column name	R-type	Explanation
variable	character vector	Variable names
distribution	character vector	Marginal distribution types
lower	numeric vector	Marginal 5%-quantiles
upper	numeric vector	Marginal 95%-quantiles

Frequent optional columns are:

Column name	R-type	Explanation
description	character	Short description of the variable.
median	cf. estimate	Marginal 50%-quantiles
method	character vector	Methods for calculation of marginal distribution parameters

Columns without names are ignored. Rows where the variable field is empty are also dropped.

The structure of the correlation file (optional)

File name structure: <marginal-filename>_cor.csv

Columns and rows are named by the corresponding variables. Only those variables need to be present which are correlated with others.

The element ["rowname","columnname"] contains the correlation between the variables rowname and columnname. Uncorrelated elements have to be set to 0. The diagonal element ["name","name"] has to be set to 1.

The matrix must be given in symmetric form.

Deprecated input standard (estimate_read_csv_old)

File name structure of the correlation file: <marginal-filename>.csv_correlations.csv

Value

An object of type estimate which element $marginal is read from file fileName and which element $correlation_matrix is read from file gsub(".csv","_cor.csv",fileName).

See Also

estimate_write_csv, read.table, estimate

estimate_read_csv, read.table, estimate

Examples

 # Read the joint estimate information for the variables "sales", "productprice" and 
 # "costprice" from file:
 ## Get the path to the file with the marginal information:
 marginalFilePath=system.file("extdata","profit-4.csv",package="decisionSupport")
 ## Read the marginal information from file "profit-4.csv" and print it to the screen as
 ## illustration:
 read.csv(marginalFilePath, strip.white=TRUE)
 ## Read the correlation information from file "profit-4_cor.csv" and print it to the screen as
 ## illustration: 
 read.csv(gsub(".csv","_cor.csv",marginalFilePath), row.names=1)
 ## Now read marginal and correlation file straight into an estimate:
 parameterEstimate<-estimate_read_csv(fileName=marginalFilePath)
 print(parameterEstimate)

decisionSupport documentation built on Oct. 6, 2023, 1:06 a.m.