search

ropensci / EML

Ecological Metadata Language interface for R: synthesis and integration of heterogenous data
https://docs.ropensci.org/EML
Other
97 stars 33 forks source link

[Units] Discussion about current unit list #343

Open earnaud opened 2 years ago

earnaud commented 2 years ago

Hi ropensci team,

I am working again on some units issue on my own work, and I have some questions/ideas. EDIT: I deleted the two previous issues #330 and #314 that I had previously opened to merge them into a discussion here.

Current unit list structure

get_unitList()$units returns a very useful data.frame of all available units. However, some seems to be duplicated and others do not have unitType where they shall have one (spectral radiance for example). I have already worked on those to fill the gaps in the "unitType" column. You can find the file attached (yes, .ods but I wanted to keep the formatting ok). The bold text is the one I reworked. Let me know if this is a convenient document for you. I have a doubt about "PhotosyntheticPhotonFlux" as unitType, since it is more the name of a measure and not a proper type.

EML_units.ods

Missing units

Some colleagues did not find some units in your unit list. However, these units (like Julian_year) are listed in the UniData documentation (https://www.unidata.ucar.edu/software/udunits/udunits-2.2.28/udunits2-common.xml). Since you refer to unidata in the unitList, I wondered what choices prevented you from using these units too. Also, I am willing to review the unit List and add the missing units I noticed. In the same way, I have users working in paleontology, and some of their time periods are expressed in thousand of years (or even millions): would you think this would be okay to add such units as "kiloyears" (or equivalent) without interfering with the EML standard unit list? (I believe not but there might be a way I did not think about). EDIT: I also saw that parts per million/billion/trillion/volume were missing. All are not crucial but ppm and ppb are regularly used.

Thanks again for your work, and let me know if I can help on this part of the tool, I am more than willing to help !

earnaud commented 2 years ago

While working on the unidata units, I observe there are no "deciBel" entry (bu ta wide set of units derived from deciBel)

earnaud commented 2 years ago

New question:

Intended use of SI

Is the parentSI meant to be used with only the seven base SI units? Or are SI-derived units valid in this column?

earnaud commented 2 years ago

Well, here it is: a reviewed list of the EML::get_unitList()$units. This file is built on the Unidata unit lists (https://www.unidata.ucar.edu/software/udunits/udunits-2.2.28/udunits2.html#Database) I listed units programmatically in a table with their name, udunit name and definitions. I then removed all duplicates or marked them as deprecatedInFavorOf. Then searched for definition and conversion in SI units. Most lines have a green background, meaning I am 90% sure of the content, some have a lightgreen background meaning I would appreciate those units to be reviewed, and some have an orange background meaning I am not sure these are even units (like Pi or Avogadro's number). I also added kiloYear (kyr) ad megaYear (myr) since these are commonly used by paleontologists.

EML_files_2.ods

earnaud commented 2 years ago

DeciBel and logarithmitc units

Yes, another topic: how could deciBel be handled in the {EML} way, since it is defined as a logarithmic value?

earnaud commented 2 years ago

Arf, just got a view into the code and saw this issue shall get into https://github.com/ropensci/emld/issues (and more broadly into EML standard)