Add units and time-interval (avg,sample) to point output about chm HOT 6 CLOSED

Units in the variable name?

from chm.

I was thinking as separate header lines like:

variable1, variable2, variablen
unit1, unit2, unitn
avg1, sample2, avgn,
data, data, data,
data, data, data

from chm.

Gross to post process though?

from chm.

Depends on your definition of gross :). I think it is easy to import in pandas. I do this for the station files here that follow that format.

# Ascii data format info
c_header = 4 # Header lines
c_column_line = 1 # line where column names start
c_delimiter = ','

# Import data to pandas dataframe
dat = ascii.read(cfile,header_start=c_column_line,data_start=c_header,delimiter=c_delimiter)
datain = pd.DataFrame(dat.as_array())

# Import header info
headerinfo = pd.read_csv(cfile,nrows=2,skiprows=1)

Having received many data sets without units/metadata in the past, I am a big proponent of including metadata. Going to a hierarchical data format like netcdf might be easier to include this extra meta data.

from chm.

I've had some which have almost unparsable headers due to a mix of '_', ' ', etc. As long as the format is well defined, I have no problem doing this.

I've always wanted the models to declare their outputs in a manner that enforced units between models. In fact, I've actually had bugs where I messed up the units. Something like this
http://www.boost.org/doc/libs/1_61_0/doc/html/boost_units/Examples.html

would be ideal, I think.

from chm.

Closing to split into two issues

from chm.