#!/bin/sh
# filename: try_hour_avg
# input: 2 columns: decimal.day, H2O2 air 
# output: averages of the above columns for every hour
# doesn't report hourly average for missing hours 

# ddy       H2O2
#159.538194	0.58
#159.565972	0.58
#159.593750	0.94
#159.621528	0.49
#159.649306	0.70
#160.461806	0.60
#160.496528	0.63
#160.524306	0.68
#160.552083	0.47
#160.579861	0.66
#160.607639	0.58
#160.635417	0.42
#160.670139	0.55


# first do some command-line checking
if [ $# -ne 1 ]; then printf "usage: try_hour_avg file\n"; exit; fi
if [ ! -f $1 ]; then printf "file: $1 does not exist\n"; exit; fi

# prompt for output filename
printf "\nPlease enter output filename:\n"
read outname

cat $1 | nawk '

BEGIN{ printf("%-12s %-12s %-22s\n", "Day_of_year", "H2O2 [ppbv]", "Number of data points") }

{ day[NR] = $1; value[NR] = $2 } # get hour and values, save in arrays

END{
fh = 1/24  # fh = fraction hour = 0.04167 

i = 1
frac = day[i] % 1   # get only the fractional part, e.g. 0.538194
inc = int(frac / fh)
sh = int(day[i]) + (inc * fh)  # sh = start hour 

while ( i < NR ) {
	while ( day[i] < (sh + fh) && i <= NR) { 
		sum += value[i] 
		count++
		i++
	} 

	printf("%-12f %-12f %-22d\n", sh, sum/count, count)
	sum = 0
	count = 0
	frac = day[i] % 1
	inc = int(frac / fh)
	sh = int(day[i]) + (inc * fh) 
}

}' > $outname # end END

printf "\noutput stored as file: $outname\n"