Cull Help


Back to Index

Cull Data:

CULL DATA refers to the process of removing spurious or unwanted points from a dataset.

The program initially reads in all data; data which do not pass the cull criteria are identified as bad data and are removed. Only those data which pass the cull crteria are then passed on to the SMOOTHING and DECIMATING and plotting routines.

Once the data input and datatype files have been selected, a user can specify culling criteria for removing bad (unwanted) data from the raw input files by pressing the CULL DATA... button in the OPEN DATA... window. The user can also specify removal of all data values within nn cm of the top of a core (the uppermost 5-10 cm of a core are commonly disturbed during drilling and core handling).

If a dataset from the same site has been read previously, culled using the cull parameters, and the CULL TABLE was saved (Sitexxxx.cull), then that CULL TABLE can be applied when subsequently reading in another dataset from the same site (see below) by selecting the "Use existing Cull File..." option in the CULL DATA window and selecting the appropriate CULL TABLE (Sitexxxx.cull).

If no culling of data is desired, the user selects the NO CULLING radio button and closes the window - note that the CULL DATA... button label then changes to NO CULLING... to indicate the change in culling status.

Culling data: To cull data, the user selects the series of logical expressions which should be used to filter out bad data values. One remove unwanted values by selecting the appropriate data filter with the (<=, <, =, =>, >) ; (AND, OR, ONLY) ; (<=, <, =, =>, >) logical expression buttons.

For example, if one wanted to remove all GRAPE density values in the dataset which were less than 1.03 g/cm3 (seawater) or greater than 2.8 g/cm3 (calcite), the following culling criteria would be applied:

Cull data values < 1.03 "OR" > 2.80

Note that the logical "OR" is selected instead of "AND" because one is removing data values which are either less than 1.03 or greater than 2.80 g/cm3.

If one just wanted to cull data values less than 1.03 (i.e. core voids) the following criteria would be applied:

Cull data values < 1.03 "ONLY"

Culling P-Wave data: P-wave velocity data in the "MST95" format have a special, additional culling feature based on the "Signal Strength" values (0-256), which provide a digital measure of the quality of the P-wave velocity determination based on transducer contact and signal amplitude. P-wave data can be culled effectively by setting the "Signal Strength" criterion to values >= 160. The P-wave data can be culled additionally using the logical expression buttons.

---------

CULL TABLE

After data have been read in and certain cull parameters have been applied, the program keeps track of all data which have removed and the reason why they were removed. These data are stored in a binary file called a CULL TABLE which lists the values which were culled and a digital identifier which indicates the reason for culling (Hash Table).

The reason for saving a CULL TABLE is that they can be applied to identify core voids and other coring problems within any dataset. For example, one can initially cull all GRAPE density values less than 1.03 to define core voids. The resulting CULL TABLE can then be used to identify and remove core voids from other volume-dependent MST measurements such as magnetic susceptibility and gamma ray activity.

____

updated: 30 April, 1996


Back to Index

Send comments on the help to Html conversion to:

Bill Kamp