Baseline transformations involve specifying sort variables. Sort variables define unique profiles for which the change from baseline is computed separately. Typically, a single variable representing subject IDs is sufficient. Other data may require more than one variable, such as Subject and Period. The baseline transformation sorts the output worksheet in alphabetical or numerical order, based on the sort variables.
Rules for datasets used in a baseline transformation:
By default, Phoenix uses the value corresponding to time=0 as the baseline for each unique profile. If a profile has more than one value at time zero, Phoenix uses the first value as the baseline.
If a zero time value is not in the dataset, either the baseline value (normally the lowest time) must be specified (use the Fixed Value field) or the minimum time for each profile can be used (select MinimumPerGroup as the Starting Point) to generate the baseline transformations.
If the column mapped to x Column has a zero or missing value at the time specified for the baseline, then the % Change from Baseline and Ratio from Baseline options will produce empty results. An alternate baseline time value or the MinimumPerGroup option may be used.
From the Transformation menu, select the type of baseline transformation.
% Change from Baseline calculates the percent change between each value and the baseline value.
result=(x – baseline)/baseline*100
Change from Baseline calculates the change between each value and the baseline value.
result=x – baseline
Ratio from Baseline determines the ratio to the baseline value.
result=x/baseline
In the Main Mappings panel, map the input data to the x Column, Sort, and Time contexts.
Choose the Starting Point.
MinimumPerGroup will use the lowest time point in each profile to generate the baseline transformation. This option means that the baseline transformation uses a different starting point for each profile that is defined by the sort key(s).
Fixed Point uses the starting time value entered in the Fixed Value field.
If a time column is in Text format in the input dataset, then the fixed value entered in the Fixed Value field must be in the same format as in the dataset (e.g., if the first time point used as the baseline is 1:00, then 1:00 must be typed in the field). For acceptable dates and times in Text format, see “Date and time formats”.
If the Fixed Value is higher than the highest time value, then no output values are produced in the new column.
From the Treat Time Values as menu, choose whether transformed time column format is numeric (Numeric Format) or text format (Date\Time Format).
When Date\Time Format is selected, use the Time Units menu to select the units for the transformed time column.
If this unit does not match the input data units for the time column, the values will be converted to the selected units.
When Date\Time Format is selected, check the Create Transformed Time Column box to generate a second time column in the output based on Text time column but in numeric format. If unchecked, there will only be one time column that is in text format.
Continue with Step 3 in the “Options tab” section.
Legal Notice | Contact Certara
© Certara USA, Inc. All rights reserved.