Page 1 of 1

nearest neighbour, linear and zero fill missing data

Posted: April 1st, 2014, 8:45 am
by lnicolot
I would like to have more information about this three functions: are they related to the peak detection using the watershed algorithm?
What is the difference between nearest neighbour and linear interpolation?
In zero-filling missing data function are the smaller neighbours of each peak added without any interpolation to create the final 2D peak?

Thank you

Re: nearest neighbour, linear and zero fill missing data

Posted: April 2nd, 2014, 6:21 pm
by tjanke
These options are used to interpolate raw data at import. Data files with time stamps (scan acquisition times) for each data point will have this option enabled.

When an interpolation option is not selected, the software assumes every data point to have a uniform Sampling Rate.
When an Interpolation option is selected, the software will try to match data point times to given the Sampling Rate. If the software does not find a data point for the expected time, it will use the selected interpolation option to account for the missing value.

Linear Interpolation: Find the data points before and after the desired time, and calculate the desired value using a linear function.
Nearest Neighbor: Copy the data point with the closest time.
Zero Fill: Leave the point blank.

  • This setting is not directly used in the peak detection, but the difference in the chromatogram may have an effect.
  • This can be a used to handle data with variable sampling rates.
  • Interpolation affects the raw data.
  • Zero Fill is intended to affect the raw data the least but may create gaps.