CONFIGURE YOUR KEY VARIABLES

AutoDiscovery must be properly configured to carry out an efficient discovery process and to provide relevant results.

Consolidation variables

 

When a data file is imported in AutoDiscovery, it is automatically consolidated with the rest of data already imported.  To do that, AutoDiscovery makes use of the "consolidation variables" previously configured by you. 

 

Consolidation variables are the columns in your data files that uniquely identify each of your experiment subjects. E.g. If your experiment consists in studying the effect of a substance in a variety of patients, your consolidation variables would probably be those that identify each patient (code, name or similar). However, if a particular patient is provided with different substances then the consolidation variables might include a unique trial identification (trial number or similar).

What is really important is that a particular combination of values of the consolidation variables might not be repeated in different rows of your consolidated table. 

 

Another relevant fact is that every data file imported should contain at least one (if not all) of the variables configured as "consolidation variables". Otherwise, the data file will simply be appended to the end of the already consolidated data table.

 

When the first data files are consolidated, AutoDiscovery automatically finds the common variables and configures them as the consolidation variables. However, if you want to change it later: 

 

  1. Select the "Consolidate" tab and click the "Configure Consolidation" button.

  2. Select the consolidation variables in the list of variables on the left side of the window. If more than one variable is to be selected, keep the Ctrl key pressed as you click on them.

  3. Click the "Assign" button located under the "Data consolidated by" field within the "Key Variables" section on the right side of the window.

 

The selected variables are included in the "Data consolidated by" field.

 

To remove one consolidation variable:

 

  1. Check the box corresponding to the variable within the "Data consolidated by" field.

  2. Press the "Unassign" button to remove the variable.​

Subject grouping variables

 

The discovery process consists in finding out relevant relationships among every combination of variables (and ratios). These relationships are evaluated in every subset of your consolidated data which are obtained by filtering the samples using the different values of the qualitative variables configured within the "Subject groups" section.

 

Notice that this is not a mandatory configuration but it really helps AutoDiscovery to improve the discovery process and the results provided.

However, the more variables are configured as subject grouping the more subsets will be analysed and thus the more time will be required to complete the discovery process.

 

To configure the subject grouping variables:

 

  1. Import and consolidate all the experiment data files.

  2. Select the "Discover" tab and click the "Configure Key Variables" button.

  3. Select the key variables in the list of variables on the left side of the window. If more than one variable is to be selected, keep the Ctrl key pressed as you click on them.

  4. Click the "Assign" button located under the corresponding field within the "Subject groups defined by" section on the right side of the window.

 

To remove one subject grouping variable:

 

  1. Check the box corresponding to the variable within field.

  2. Press the "Unassign" button to remove the variable.​

Variables to explain

 

This section allows you to explicitly determine which variables will be considered during the discovery process so that only those relationships in which any of these variables participates will be evaluated.

It is important to remark that if this section is left empty, all the variables will be considered so it'd be a good idea to select some of them if your discovery process is expected to take a long time.

 

Again, the more variables are configured to explain, the more relationships will be analysed and thus the more time will be required to complete the discovery process.

 

To configure the list of variables to explain:

 

  1. Import and consolidate all the experiment data files.

  2. Select the "Discover" tab and click the "Configure Key Variables" button.

  3. Select the variables to explain in the list of variables on the left side of the window. If more than one variable is to be selected, keep the Ctrl key pressed as you click on them.

  4. Click the "Assign" button located under the corresponding field within the "Variables to explain" section on the right side of the window.

 

To remove one subject grouping variable:

 

  1. Check the box corresponding to the variable within field.

  2. Press the "Unassign" button to remove the variable.​

Variables to ignore in calculations

 

Irrelevant variables consolidated in your experiment project (e.g. duplicated information, numerical variables that must not be considered in correlations, etc.) can be ignored during the discovery process calculations. 

To configure the variables to ignore in calculations:

 

  1. Import and consolidate all the experiment data files.

  2. Click the AutoDiscovery's title bar to drop down the menu and click the "Configure Variables" button.

  3. Select the variables to ignore in the list of variables on the left side of the window. If more than one variable is to be selected, keep the Ctrl key pressed as you click on them.

  4. Click the "Assign" button located under the "Variables to ignore in calculations" field within the "Key Variables" section on the right side of the window.

 

The selected variables are included in the "Data consolidated by" field.

 

To remove one consolidation variable:

 

  1. Check the box corresponding to the variable within the "Data consolidated by" field.

  2. Press the "Unassign" button to remove the variable.​