Samples can be of many different types, such as a core, a liquid sample, or even the hole itself. All samples are related to other samples by parentage (e.g., cores are children of holes, sections are children of cores, section halves are children of sections, etc.). Many layers of parentage can exist down to the bottom of a given hierarchy tree. Holes are at the top of each parentage tree, so in the database they have a default “parent” sample_number = 0. (LIME cannot assign samples to this default parent.) Site and expedition information are metadata about the hole; these two parameters are not part of the parentage.
In LIME, operations on samples must take the parentage hierarchy into account, and in many cases must follow the parentage tree to the bottom of each and every branch. This can be true if a sample is canceled, edited, or moved from one part of a parentage tree to another (reparenting).
For example, if a section is canceled using LIME, the program cancels the section, any tests and results run on the section, the section halves, any tests and results run on the section halves, and so on, down to the last toothpick sample taken from that section, and its tests and results. Likewise, if a thin section billet is reparented (i.e., moved) to another section because it was created in error, then the billet’s label information changes as well as does the label information on the thin sections made from that billet.
LIME is designed to take parentage into account, but the user must be aware of the consequences of changing parentage in order to utilize LIME properly. Reparenting the billet in the above example automatically moves the thin section parentage as well.
A collection of results is organized under a descriptive wrapper called a test. Test records are defined by the ANALYSIS table and populated once assigned to a sample. Each test in the TEST table is associated with one and only one sample, and each test has one or more associated results. Therefore, each test has a “parent” (its sample) and one or more “children” (its result or results). Working with tests is in general much simpler than working with samples because this parentage structure is always the same and does not descend any further than the RESULT table.
Another way of explaining this is such: a result is the smallest unit in the parentage structure. A result is produced by performing a test. There can be multiple results for a given test if the test was repeated. A test is always associated with an analysis. An analysis can be run on different samples. An example of a test is CHNS run on a powder sample, or a set of MS data run at different offsets on a section.
Canceling a test record using LIME also cancels its result records—because the results are children of the test.
Individual results are stored in the RESULT table. These results are grouped under their parent tests and tied to samples through the TEST table, as explained above (i.e., results are not directly associated with samples in LIMS). Each result is associated with only one test; however, many results and replicate results may be associated with that same test. Results have a “parent” (the test record); they never have “children.”
Some results are collected in replicate sets, such as the results from most of the core loggers. For an MS test, the results might include magnetic susceptibility, instrument serial number, and offset on the section. The first measurement on a section is by definition replicate /0; the second is replicate /1; the third /2; etc. For measurements using replicates, some of the results are not repeated except on replicate /0 if they apply to all measurements. In the example below, the first six lines are the common results and are stored along with the 0.25 cm offset results.
user=houpt text_id=SECT2163391 instrument=MSLP_476 instrument_group=FAST_TRACK observed_length=1.5065 comment=
offset= 0.25,magnetic_susceptibility = 93.0 offset= 5.26,magnetic_susceptibility = 109.0 offset= 10.27,magnetic_susceptibility = 74.0 offset= 15.28,magnetic_susceptibility = 80.0 offset= 20.29,magnetic_susceptibility = 83.0 offset= 25.30,magnetic_susceptibility = 89.5
|
Figure 2. Example Replicate Results.
Individual results cannot be canceled using LIME. To cancel a result the user must cancel the parent test and all results associated with that test. LIME does not allow editing of individual results (with the single exception of a container number for the MAD analysis).