Scaling multiple columns
Hi,
I'm trying to normalize a large data set to remove errors caused by equipment. I have 101 columns (samples) that are in time order by column. These columns include multiple quality control samples (1 in every 10 or so) that should have all close to equal values. These however go down linearly and therefore all the data is transformed down (seen in replicate samples place randomly across the running order). The first and last sample are QCs. Is there a way to scale up all the values so that the QCs would be ~equal and my data would be transformed?
e. Just to clarify what I would like to do, is to use the first QC as a pivot point and the last QC as the lever to transform values in between.
Last edited by Arto Vesterbacka : November 16th 12 at 07:52 AM
Reason: Added more information about what I want to do
|