panels/E2_NumericVariables/4_transform.variables.help.md

In iNZightVIT/Lite: iNZight Lite

How to use the transform panel?

1. Select one or more columns:Using the dropdown menu at the left hand side labeled "Select Columns" select any number of columns available. The selected columns should be displayed in the middle of the screen.
2. Select transformation type:Select any (one) of the transformation types in the dropdown menu labeled "Select transformation:". After the selection, an additional number of columns is added to the presented data.
3. Transform columns:Step one and two only presented the data and help with the selection. Until now no additional data was added. To add the additional columns to your data, press the "Transform" button. After that the additional transformed columns are added and will be available.

Add Transformation

Adds up all selected columns which are numeric.

Subtract Transformation

Subtracts all selected columns which are numeric such that "A-B-C". Where A, B and C are the selected columns.

Multiply Transformation

Multiplies all selected columns which are numeric.

Divide Transformation

Divides all selected columns which are numeric.

Log Transformation

Performs a log transformation of all selected columns. By default this is the natural log. For more information please look at the R function "log".

Root Transformation

If the columns are of class numeric the square root of each value is estimated and returned.

Square Transformation

All columns of class type numeric are squared, such that XX.

Abs Transformation

Of all numeric columns from the selected column the absolute value is returned. See the "abs" function in R for more information.

Center Transformation

All columns (character, factor and numeric) are centered to mean 0. Such that from column X the mean of X is subtracted from X. Character and factor values are converted to numeric such that whith the order of the factors values from 1 to the nuber of unique factors are asigned to each value in the column.

Standardize Transformation

In a first step exactly the same method is performed as described in "Center Transformation" above. As an additional step the result from the first step is divided by the standard deviation of the column.

Median Split Transformation

All numeric selected columns are converted into a binary variable such that all values above the median are classified as "high" and all values below the median are classified as "low".

Reverse-coding Transformation

The selected numeric columns are revers coded such that the minimum and maximum of Column X is added and X is subtracted from this value ((min(X)+max(X))-X).

Copy Transformation

Copies the selected coulms so that they become duplicates in the current data set.

Change sign Transformation

Performs the action X-1 such that all negative values become positive and all positive values become negative.

Convert to categorical Transformation

Changes any variable/column to a factor value. Note, all character columns are already treated as factor.

iNZightVIT/Lite documentation built on April 13, 2024, 8:03 p.m.