MissingDataGUI (version 0.2-5)

WatchMissingValues: The Main Window of Missing Data GUI.

Description

This function is to open the missing data GUI. The widgets shown in the GUI include: a table of all variables in the dataset, a checkbox group of categorical variables to condition on, a table of variables which have missing values to coloy by, a radio of imputation methods, a radio of graph types, three command buttons, and a graphics device. In this GUI the user can: 1)change the name and class of the selected variable; 2)look at the numeric summary for the missing values in the selected variables; 3)look at the plot of imputed data, under one of the imputation methods and one of the graph types and one color-by variable, with or without the conditions; 4)export the imputed data as well as the missing shadow matrix, and save them to a data file(csv).

Usage

WatchMissingValues(h, data = NULL, gt = NULL, size.width = 1000, size.height = 750,  ...)

Arguments

h

A list with components obj referring to the button "Watch Missing Values" in MissingDataGUI.

data

A data frame which is shown in the missing-data GUI. If it is null, then parameter gt must not be null.

gt

A widget created by gtable(). It should be passed from MissingDataGUI.

size.width

the width of window. Default to be 1000, and the minimal is 800.

size.height

the height of window. Default to be 750, and the minimal is 600.

...

Other parameters to be passed to this function.

Details

The missing data GUI consists of two tabs. In the summary tab, there are a list of all variables, a list of variables having missing values to color by, two radios for imputation methods and graph types respectively, a checkbox group for the conditional variables, four buttons and a graphics device. In the help tab, the layout is the same as the summary tab. But when the users move their mouse on those widgets, or click any of those radios or buttons, the functions of all widgets will be described at the place of the graphics device. The attributes of the variables can be changed. If the user double clicks on any variables in the top left table of missing-data GUI, an attribute window will pop up. Then the name could be edited, and the class could be changed to one of the four classes: integer, numeric, factor, and character. When a numeric variable is changed to a categorical variable, the condtions in the bottom left checkbox group will be updated. If the list of the color by variables is very long, the selector allows text entry to find the variable when this widget is active.

Examples

if (interactive()) {
    data(tao)
    WatchMissingValues(data = tao)
    data(brfss)
    WatchMissingValues(data = brfss)
}