--- Sample data
 dep tt ndep tim lat lon stn dtyp in1 dpath dfile

dep - cell array - depths of each cast in a separate cell
tt  -  "     "   - temperature " " "
ndep - vector of number of depths in each cast
lat -   "
lon -   "
stn -   matrix of char station numbers
dtyp -  vector of Data Type for each station. Note that Data Type is a char*2
        object. To get it into Matlab, I have melted it into a single number
 (char(1)*256 + char(2)). See where it is used in cal_stat.m
in1 -   index of the loaded region casts out of all casts in the keys file
dpath - path to start of data files directory tree root
dfile - sample keys file
 

--- Archive data
 adep att andep atim alat alon astn adtyp in0 apath afile

Same interp as for sample, above
 

--- region plot
  axs reg

axs - axis of plot
reg - vertices of polygon defining the region
 

--- stats selection
 stnams sttype binnams bintype btreq binstr

stnams - names of the different stats
sttype - selected stats type (ie between 1 - 11)
binnams - names of binning methods
bintype - selected  " "
btreq  - vector defining which stats types require binning of data
binstr - string containing default (or modified) specification of bins for
         each stats type.
 

--- stats results
  bins cnts mns sds pctls ireg cval iscr svscr seas

bins - vector of last defined bins
cnts - counts in each bin at last calculation
mns -  means " " " "
sds -  Standard Dev " " "
pctls - 4 percentiles for each bin " "
ireg - counter to increment output file name for each different stat computed
cval - results for stats generating a single value per cast
iscr - last screening result, 1=rejected. If last calculation produced a single
       value per cast, iscr is same shape as cval. If last was t(z), iscr is
       same shape as relates to loaded temperature data. iscr not yet coded
       for other stats.
svscr - 1 if last stats were single-value-per-cast type.
seas - vector of temporal bins, defined as day-of-year
 

--- extra stats treatment choices
  prefilt nsd to2m wgt
 
prefilt - 1=> running weighted mean prefilter high res casts
nsd     - Number of Standard Deviations to use for screening
to2m    - 1=> interpolate low res casts to 2m
wgt     - pair defining relative weights to be applied to sample and archive
          data in producing stats.
 
 
 

Add: dectim and adectim plus all the parameter names (ssts,t100s t250s mld1s mld2s, etc._