# Collocation result file

The collocation result file is a simple comma separated (csv) file that provides the list of matching pairs of measurements from two datasets (dataset A and B). Both datasets can consist of multiple files. Each measurement from dataset A can match with multiple measurements of dataset B and vice-versa. This means that the combination of filename_a and measurement_id_a can occur multiple times in a collocation result file and the same holds for a combination of filename_b and measurement_id_b. Each pair is uniquely identified by a collocation index (collocation_id).

A collocation result file contains the following columns:

collocation_id

Unique id of the collocation pair. This id will correspond with the collocation_index variable inside HARP products after the products are filtered using a collocation result file.

filename_a

The filename of the original input file (i.e. source_product global attribute value) from the primary dataset.

measurement_id_a

A unique index number of the measurement within the file. This index number is based on the list of measurements from the original input file and corresponds to the index variable inside HARP products.

filename_b

The filename of the original input file (i.e. source_product global attribute value) from the secondary dataset.

measurement_id_b

A unique index number of the measurement within the file. This index number is based on the list of measurements from the original input file and corresponds to the index variable inside HARP products.

collocation criteria…

The remaining columns cover the collocation criteria that were provided to harpcollocate. For each collocation criterium the column will provide the exact distance value for the given collocated measurement pair for that criterium. The column label used for each criteria is the HARP variable name of the associate difference variable together with the unit (e.g. datetime_diff [s])