fgreg commented on a change in pull request #61: SDAP-105 DOMS matchup netcdf
and csv generation
URL:
https://github.com/apache/incubator-sdap-nexus/pull/61#discussion_r246135613
##########
File path: analysis/webservice/algorithms/doms/BaseDomsHandler.py
##########
@@ -124,41 +127,63 @@ def __packValues(csv_mem_file, results):
headers = [
# Primary
- "id", "source", "lon", "lat", "time", "platform",
"sea_water_salinity_depth", "sea_water_salinity",
- "sea_water_temperature_depth", "sea_water_temperature",
"wind_speed", "wind_direction", "wind_u", "wind_v",
+ "id", "source", "lon (degrees_east)", "lat (degrees_north)",
"time", "platform",
+ "sea_surface_salinity (1e-3)", "sea_surface_temperature
(degree_C)", "wind_speed (m s-1)", "wind_direction",
+ "wind_u (m s-1)", "wind_v (m s-1)",
# Match
- "id", "source", "lon", "lat", "time", "platform",
"sea_water_salinity_depth", "sea_water_salinity",
- "sea_water_temperature_depth", "sea_water_temperature",
"wind_speed", "wind_direction", "wind_u", "wind_v"
+ "id", "source", "lon (degrees_east)", "lat (degrees_north)",
"time", "platform",
+ "depth (m)", "sea_water_salinity (1e-3)",
+ "sea_water_temperature (degree_C)", "wind_speed (m s-1)",
+ "wind_direction", "wind_u (m s-1)", "wind_v (m s-1)"
]
writer.writerow(headers)
+ #
+ # Compress multiple depth variables into one depth array
+ #
+ depthAggregate = []
Review comment:
I think this needs further discussion. I know ICOADS stores two separate
depth fields (one for sst and one for sss) per record in their index. I'm not
sure about FSU but this is why we agreed to have two fields for depth in edge.
I think by trying to combine it into one value we are making a decision for the
in situ provider when we should just be passing along the data that was given
to us.
Same for filling missing data with 0; that implies that the in situ provider
said that the data was recorded with a depth of 0. For example the FSU
translation specification states "Note that a SAMOS value of -9999 or -8888 is
interpreted as “missing” or “special” and is left out of DOMS." for
Sea_water_temperature_depth. In this case we might end up reporting a 0 depth
when that is in fact incorrect.
IMO we need to retain two depth fields: Sea_water_temperature_depth and
Sea_water_salinity_depth
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
