Laser logs and mass spec sample metadata speed up the process of creating selections in iolite considerably. However, sometimes we want to further refine those selections for a variety of reasons. For example, maybe the laser log and mass spec don’t quite agree on the timing of things, or you want to automatically avoid parts of the signal where you’ve ablated through the target. Read on to find out more about iolite’s new selection refiner tool!
Overview of the user interface
The selection refinement tool (included in iolite 4.5.2+, some features described in soon to be release v 4.5.3) can be accessed via the Tools → Refine selections menu item. A dialog similar to below will appear:
In this dialog there are a few different areas of note:
- The Selection Groups area on the left. In this list, you select the group(s) that you would like to apply the refinement strategy to.
- The Strategy area comprising the bulk of the window. In this area, you select the strategy type at the top and the settings for that strategy are shown below.
- The Output area at the bottom. In this area, you specify the suffix that will be used when refining the selections in a group. For example, if you have selected a group called ‘Unknowns’ and the suffix is ' refined’, a new group will be created called ‘Unknowns refined’ that contains the selections of ‘Unknowns’ that have been modified by the selected strategy.
Refinement strategies
There are a number of refinement strategies included with this tool and you can also add your own since this plugin is python-based. The included strategies are briefly described below.
Criteria
This strategy is similar to ‘automatic selections from channels’. It finds the longest segment of each selection that matches a list of criteria. For example, you could use this strategy to adjust a selection to avoid inclusions or parts of a zircon analysis with high 204Pb.
Mean shift clustering
This strategy uses mean shift clustering on the data for the specified channels during each selection to find the cluster in the selection. For example, you could use this strategy to adjust a selection to the dominant phase of an analysis.
Rolling standard deviation and error
This strategy searches the selection for the segment with the lowest standard deviation or standard error for the selected channel. For example, you could use this strategy to find part of a selection that has the most homogeneous 207Pb/206Pb as shown below where the full selection (red ellipse) was refined based on 207/206 (blue ellipse).
Extremes
This strategy looks for the segment (with a minimum length criteria) that minimizes or maximizes the selected channel on average. For example, you could use this strategy to adjust a selection to match the highest (perhaps to avoid drill through) or lowest (perhaps to minimize discordance) segment. Shown below is the same example as above but instead refining the selection to minimize discordance (resulting in a very similar sub-segment).
Synchronize
This strategy adjusts the selection such that it is optimally synchronized with the data within some search window specified as the maximum time difference. For example, you might use this to correct small differences in timing between the recorded mass spec data and the laser log.
The end
That was a quick look at the new selection refinement tool in iolite 4. If you have any questions, comments or suggestions about this Note, you can discuss it here.
