search

tristan-sebens / metl

An R package developed by and for the AFSC MESA group for extracting and standardizing data from telemetry tags.
GNU General Public License v3.0
0 stars 0 forks source link

Writing data to `.csv` files #76

Closed CindyTribuzio-NOAA closed 2 days ago

CindyTribuzio-NOAA commented 1 month ago

Can file names be more descriptive? Please include tag_type and tag_num in the file name

What is the timestamp in the instant.csv file? Can this be converted to a meaningful value?

There was no summary.csv file for the test tag (ET5728). Can there be some output text that explains why? Or if there are errors.

tristan-sebens commented 1 month ago

The TIMESTAMP field is a POSIXct timestamp number. This is an advantageous data format because it can specify an instant in time without having to specify a timezone. However, it does have the disadvantage of being completely unreadable to the user.

In the development version of the package (not released yet) the package now outputs a FIELD_METADATA frame, which describes the contents and units of all of the output fields. This will communicate to the user that the TIMESTAMP field is a POSIXct timestamp, which will allow them to convert it into a character timestamp in the timezone of their choice.

However, in the version you are currently using, this output doesn't exist, so there is no documentation about the timestamp format.

I'll add a note to the README

tristan-sebens commented 1 month ago

Character timestamps have now been added to the output FieldMap objects, so the POSIXct timestamps will be accompanied by human-readable timestamps, which will automatically be formatted to the timezone of the system decoding the tag data (#67).

tristan-sebens commented 1 month ago

To your third point: Not all tags produce all data types, so the summary dataset will only be available for tags which produce summary data.

I've added a table to the README detailing which tags produce which types of data.

tristan-sebens commented 2 days ago

Due to current implementation architecture, it is actually somewhat difficult for the code which saves the output data to know what are the id and tag type values of the current tag.

For now I would recommend simply naming the output directory with the tag id and tag type values. However, I will add this feature to the backlog for potential future implementation.

Feature request recorded in issue #113