analyse.py

import csv
import numpy


with open("c46145.csv", "rb") as f:
    r = csv.reader(f)

    headers = r.next()
    print headers
    vwh_idx = headers.index('VWH$')
    wspd_idx = headers.index('WSPD')
    gspd_idx = headers.index('GSPD')
    sstp_idx = headers.index('SSTP')

    data = [(float(row[vwh_idx]),
            float(row[wspd_idx]),
            float(row[gspd_idx]),
            float(row[sstp_idx]))
            for row in r if row[vwh_idx] is not None and row[vwh_idx] != '']

    data_headers = ['Significant Wave Height',
                    'Wind speed',
                    'Gust speed',
                    'Sea Surface Temperature']

    data_with_headers = zip(data_headers, zip(*data))
    # zip(*data) means unzip(data)

    for name, dim in data_with_headers:
        print name
        print "Median: {}".format(numpy.median(dim))
        print "Mean: {}".format(numpy.mean(dim))
        print "Std: {}".format(numpy.std(dim))
        print "---"

    # Model response conditions
    total = 0
    no_response_possible = 0

    for vwh, wspd, _, sstp in data:
        if sstp < -3.:
            continue  # Assume nobody's operating in these conditions

        total += 1
        if vwh > 3.:
            no_response_possible += 1
        elif vwh > 1. and wspd > 10.:
            no_response_possible += 1
        elif wspd > 15.:
            no_response_possible += 1

    print "Total data points with sst > -3 C:", total
    print "No response possible in", no_response_possible, "of them"
    print "({0:.1f}%)".format(float(no_response_possible) / float(total) * 100)