Preparation of the databases

Paleoceanographer accepts calibration and sampling databases as rectangular matrices in CSV format. In the calibration database, the first column is reserved for core-top sample identifiers. The following columns correspond to taxonomic variables, where the values represent the frequencies of the taxa within each assemblage. The last six columns are reserved for the variables to be estimated. Among these, the first three correspond to the oceanographic variables used in Autoevaluation or Parameters functions (SST, seasonality, and SSS). The next two columns are used in the Analogs function to store distances and geographic coordinates, and the final column corresponds to the transformed latitude, useful for representing core-tops or sites on Mercator projection maps.

The sampling database must also reserve the first column for sample identifiers and include the same taxonomic variables, sorted in the same order as in the calibration database. However, the sampling dataset excludes the columns corresponding to oceanographic and geographic variables.

When using an alternative calibration dataset instead of PaleoUma, the number of taxonomic columns may vary; however, the final six columns must be reserved for three environmental variables and three geographic coordinates, respectively.