Level Assay Channels

Use the Geochem Analysis > GeochemUtilities > Level Assay Channels option (CHLEVEL GX) to level values in the ASSAY channels based on categories or classes defined in a channel and to produce new leveled channels in the same database.

Level Assay Data dialog options

Categorization channel

Select a channel from the list of available channels. All data values corresponding to each unique item in this channel are adjusted based on statistics of that group.

If this entry is left blank, then all data is considered to be part of the same "class".

Script Parameter: CHLEVEL.CATCHAN

Leveling method

Select the statistic used for leveling. Available options are:

  • Divide by group means

  • Divide by group medians

  • Divide by group modes

  • Divide by group log averages (geometric means)

  • Z-Scores: subtract mean, divide by std. dev.

  • Z-Scores: subtract median, divide by std. dev.

  • Robust Z-Score: subtract median, divide by 1st to 3rd percentile range.

  • Peak to noise (Response Ratio) :divide by the mean of the 1st quartile, then round to the nearest integer.

Script Parameter: CHLEVEL.LEVEL

Channels to level

"All ASSAY data" – all channels that are members of the ASSAY class will be levelled.

"Displayed ASSAY data" – only displayed channels that are member of the ASSAY class will be levelled (this is the default setting).

"Select ASSAY channels from list" -- select ASSAY class channels from a two-panel selection list.

Although ALL ASSAY channels are included in the listings (including the levelled channels), the levelled channels are automatically removed from the list before processing.

Script Parameter: CHLEVEL.DATA ("ASSAY", "DISPLAYED_ASSAY" and "LIST")

Application Notes

This GX is useful for transforming data prior to doing various statistical operations.

The GX performs the following operations on each selected assay channel’s data:

  1. All data in the channel in all selected lines is read, sorted by the categorization value, and each group is analysed to determine the required leveling statistic.

  2. Data in each group is altered based on the statistic chosen, and then written back to the new leveled channel.

Dummy values remain dummies on output.

The new leveled channel name is created using the original channel name, and appending the leveling method and categorization channel name. For instance, for group mean leveling, using the category "Area", the "Cu" channel would be leveled to the new channel "Cu_Mean_Area".

Channels which have already been leveled cannot be leveled a second time, and are automatically removed if you select (for instance) "All Assay Channels" as your channel selection.

Assay channel attributes settings.

Detection limits and outlier values are removed from the new leveled assay channels, because they are no longer applicable to the rescaled data. (It is not possible to apply the leveling to the limits because different scaling factors are used for different categories in the data.)

Units are removed from those processes which involve rescaling of the data. Symbol scaling values are removed.

Log scaling preferences are removed; i.e. the default plotting display method for the leveled data is linear.