bin_taxa()

To organize count data from a zoo_df by taxanomic categories, bin_taxa() is the way to go. This is the engine behind uvp_zoo_conc(). In it’s current configuration, bin_taxa() is intended for binning by taxa into specified depth-bins. However, it’s future use should allow for binning by season, location, or other categories. Under the current configuraiton, the zooscan option will allow for alternative binning, it might just require some hard-coding.

Main Functionality:

The main argument needed to provide is a zoo_df and an argument for depth_breaks. This is a character vector of the limits for each depth bin. For example, to distinguish the epipelagic and the mesopelagic, you can set depth_breaks = c(0,200,1200). By default, bin_by() will identify each unique taxa by the $name or $object_annotation_category column and return the count of each label in each bin. Other columns can be specified with the cat_col argument if there another column to aggregate by (ex an alternative name). Additionally, while the main use of bin_taxa() is to count the number of observations in each bin, it can be used with other functions. The func and func_col can be used to specify what other columns to apply the function to. For an example look at its application in the uvp example

Use with UVP:

To use bin_taxa() with UVP data, it is best to go through uvp_zoo_conc(). However, it is possible to use bin_taxa() with a single zoo_df. One consideration with the UVP is that if there are no observations in a depth-bin you must set force_bins = TRUE. This isn’t an issue for net samples. However, it can be an issue with the UVP because it may sample a bin with nothing in it. Likely that is if bins are too small. Nonetheless, this is already accounted for in uvp_zoo_conc().

Use with Zooscan:

Currently, the zooscan configuration will simply bin based on unique sample_id columns. Thus, it is important to be considerate of what the sample_id’s indicate. To use, simply set zooscan = T and it will return counts aggregated by sample_id.