Problems arising using vector or grid scans:

Nine sequential Pinwheel.ini vector scans interpreted by the surface() function of the R fields library. Notice the outlier affect on the contours in panel D (red-coded) which is further commented on below.

Viewing a particular Pinwheel.ini vector scan protocol of ASET software, remotely, using LogMeIn from home to access my UNE ion probe rig. Notice the one outlier measurement (red line reaching to the right margin.) The 2 µm diameter tip diffusional source with 1 M CaCl₂ in 0.1% agarose enters from the right. The 2 µm tip Ca-LIX electrode enters from left.

Using the R fields library to plot a color coded points-rendering of sampling points and signal strength. The Pinwheel.ini vector scan instruction uses 6 linear scans through a central point to focus on a central feature such as a point diffusional source.

These contour plots, in scan panels D-I, show that the six diameter scans which alternate their directions around the source measure the gradient of Ca flux differently when going up the gradient than going down a gradient. This is an indication that the electrodes were not silanized as well as possible. Work harder to improve the electrode silanization. Scanning a well understood standard source allows one to be self critical about the quality of derived data. A single outlier vector is seen to contribute to panel D resulting in a red focal spot in that panel seen as the only significantly positive vector in an overall negative flux background.
The collection of many vectors of measurements will result in occasional noise phenomena that depart from the expected pattern. Vector scan plot illustrates the measured vectors of Ca-flux eminating from a point source as directed by the Pinwheel.ini vector scan ini file. A single vector is seen to go in the opposite direction than expected. Such catastrophic or outlier vectors can be ignored or culled out from the analysis, leaving the remainder of the vectors to determine the contour plot. The R fields library allows for randomly spaced data to contribute to a contour surface plot, so elimination of a single outlier is possible and does not preclude calculating a surface.
If an outlier flux vector is not eliminated it can result in a distortion of the target pattern as observed in panel D of the contour plot above. The red spot in panel d represents one outlier vector that needs to be ignored.

Return to Data Resource