Back to PLS Help
Relative Ref.Scan Onset
Posted on 08/22/08 09:56:39
Number of posts:
Dear List,
I am wondering if the best value for relative ref.scan Onset and number of reference scans are the ones which are set as default (zero and one respectively)?If not,how one can set the best values for these options?I have read the description about these two section in the user guide,but I afraid if I have not fully understood them?It would be nice if one can clarify it with an example.
Thanks in advance.
David
I am wondering if the best value for relative ref.scan Onset and number of reference scans are the ones which are set as default (zero and one respectively)?If not,how one can set the best values for these options?I have read the description about these two section in the user guide,but I afraid if I have not fully understood them?It would be nice if one can clarify it with an example.
Thanks in advance.
David
Replies:
Untitled Post
I'm Online
Posted on 08/22/08 10:24:32
Number of posts: 291
Here is an example:
Your Onset for ConditionA is: 50 100
Your Relative Ref.Scan Onset for ConditionA is: 0
Your Number of Reference Scans for ConditionA is 1
Your onset for ConditionB is: 200 300
Your Relative Ref.Scan Onset for ConditionB is: 10
Your Number of Reference Scans for ConditionB is 20
If Normalize data with ref.scans box is checked, the datamat value will be normalized by the ref.scans that you selected. i.e.:
NewValue = (OldValue - RefValue ) * 100 ./ RefValue;
where,
RefValue for ConditionA is: ScanList(50);
RefValue for ConditionB is: AverageOf( from ScanList(210) to ScanList(219) );
Note: ScanList starts from 0.
To answer your first question, whether if is the best value all depends on your data. Basically, as is described on the user guide, it is designed to prevent huge DC offset (low frequency noise). I believe that by choosing [0 1], it will have the least impact to your data, that is why I make it as the default value.
Login to reply to this topic.
