In [1]:

    

%pylab inline
from numpy.lib.recfunctions import *


    

    




Populating the interactive namespace from numpy and matplotlib

In [3]:

    

def filternan(a):
    return a[isnan(a) != True]


    

In [15]:

    

segments = genfromtxt("segments.csv", delimiter=',', names=True)
names = genfromtxt("names.csv", delimiter=',', names=True)
callsigns = genfromtxt("callsigns.csv", delimiter=',', names=True)
imos = genfromtxt("imos.csv", delimiter=',', names=True)
fishing = genfromtxt("../Combinedfishing_2014.csv", delimiter=',', names=True)
fishing = append_fields(fishing, 'fishing', [], dtypes='<f4')
fishing['fishing'] = 1.0
tileset = genfromtxt("pipeline_205_08_07_13_33_31_mmsi_list.csv", delimiter=',', names=True)
tileset = append_fields(tileset, 'tileset', [], dtypes='<f4')
tileset['tileset'] = 1.0

d = join_by('mmsi', segments, names, jointype='outer', asrecarray=True)
d = join_by('mmsi', d, imos, jointype='outer', asrecarray=True)
d = join_by('mmsi', d, callsigns, jointype='outer', asrecarray=True)
d = join_by('mmsi', d, fishing, jointype='outer', asrecarray=True)
d = join_by('mmsi', d, tileset, jointype='outer', asrecarray=True)

d = d.filled(nan)
savez("mmsi_counts.npz", d)


    

In [30]:

    

d = load("mmsi_counts.npz")['arr_0']


    

In [31]:

    

f = d[d['fishing'] == 1]

histfig = figure(figsize=(20,5))
hist(filternan(f['seg_count']), bins=50, normed=False, color='b', alpha=0.5, log=True, label="segment count")
legend(); show()

histfig = figure(figsize=(20,5))
hist(filternan(f['radius']), bins=200, normed=False, color='r', alpha=0.5, log=True, label="radius")
legend(); show()

histfig = figure(figsize=(20,5))
hist(filternan(f['name_count']), bins=200, normed=False, color='b', alpha=0.5, log=True, label="name count")
legend(); show()

histfig = figure(figsize=(20,5))
hist(filternan(f['imo_count']), bins=200, normed=False, color='b', alpha=0.5, log=True, label="imo count")
legend(); show()

histfig = figure(figsize=(20,5))
hist(filternan(f['callsign_count']), bins=200, normed=False, color='b', alpha=0.5, log=True, label="callsign count")
legend(); show()


    

In [69]:

    

def compare(col1, col2):
    col1_coord = unique(filternan(d[col1]))
    col1_coord.sort()
    col2_coord = unique(filternan(d[col2]))
    col2_coord.sort()
    
    all_counts = zeros((col1_coord.shape[0], col2_coord.shape[0]))
    for i1, c1 in enumerate(col1_coord):
        for i2, c2 in enumerate(col2_coord):
            v = d[d[col1] > c1]
            v = v[v[col2] > c2]
            all_counts[i1,i2] = v.shape[0]

    return col1_coord, col2_coord, all_counts


    

In [87]:

    

c1, c2, p = compare('seg_count', 'name_count')


    

In [86]:

    

x = log(p)
imshow(x); show()
imshow(x[:20,:20]); show()
imshow(x[:10,:10]); show()


    

    




/usr/local/lib/python2.7/dist-packages/ipykernel/__main__.py:1: RuntimeWarning: divide by zero encountered in log
  if __name__ == '__main__':






    













    













    

In [48]:

    

f = d[d['tileset'] == 1]
print "#### 0.2"
print f[f['seg_count'] == 2.0][200]['mmsi']
print f[f['seg_count'] == 2.0][201]['mmsi']
print f[f['seg_count'] == 2.0][202]['mmsi']
print f[f['seg_count'] == 2.0][203]['mmsi']

print "#### 0.5"
print f[f['seg_count'] == 5.0][200]['mmsi']
print f[f['seg_count'] == 5.0][201]['mmsi']
print f[f['seg_count'] == 5.0][202]['mmsi']
print f[f['seg_count'] == 5.0][203]['mmsi']


    

    




#### 0.2
205336000.0
205346000.0
205482090.0
205532000.0
#### 0.5
440071000.0
440107870.0
440111111.0
440120530.0

In [49]:

    

f = d[d['tileset'] == 1]
print "#### 0.2"
print f[f['name_count'] == 2.0][200]['mmsi']
print f[f['name_count'] == 2.0][201]['mmsi']
print f[f['name_count'] == 2.0][202]['mmsi']
print f[f['name_count'] == 2.0][203]['mmsi']

print "#### 0.5"
print f[f['name_count'] == 5.0][200]['mmsi']
print f[f['name_count'] == 5.0][201]['mmsi']
print f[f['name_count'] == 5.0][202]['mmsi']
print f[f['name_count'] == 5.0][203]['mmsi']


    

    




#### 0.2
211495000.0
211501000.0
211502000.0
211516530.0
#### 0.5
261000660.0
261003220.0
261003640.0
261004780.0

In [ ]: