def test_spikes( self ):
"""
Make a simple case with spikes offset on each trace.
"""
test_data = np.zeros( (100,100) )
index = np.arange( 0,100 )
test_data[index,index] = 1.0
maxlag = 5
dip = dipsteer( test_data, 1,1,maxlag )
# Output should be all -1 except for the edge effects
same = np.allclose( dip[ maxlag:-maxlag, maxlag:-maxlag],
-test_data[maxlag:-maxlag,maxlag:-maxlag],
.01 )
self.assertTrue( same )
# Answer should not change if we increase the stepout
stepout=4
maxlag=1
dip = dipsteer( test_data, 1,stepout,maxlag )
diff = ( dip[ maxlag*stepout:100-(maxlag*stepout),
stepout * maxlag :100-(stepout*maxlag)] +
test_data[ maxlag * stepout :100-(maxlag*
stepout) ,
maxlag*stepout:100-(stepout*maxlag) ] )
diff = diff[1:-1,1:-1]
# Output should be all -1 except for the edge effects
same = np.allclose( diff, np.zeros(diff.shape ), .01 )
self.assertTrue( same )
def test_spikes(self):
"""
Make a simple case with spikes offset on each trace.
"""
test_data = np.zeros((100, 100))
index = np.arange(0, 100)
test_data[index, index] = 1.0
maxlag = 5
dip = dipsteer(test_data, 1, 1, maxlag)
# Output should be all -1 except for the edge effects
same = np.allclose(dip[maxlag:-maxlag, maxlag:-maxlag],
-test_data[maxlag:-maxlag, maxlag:-maxlag], .01)
self.assertTrue(same)
# Answer should not change it we increase the stepout
stepout = 4
maxlag = 1
dip = dipsteer(test_data, 1, stepout, maxlag)
diff = (dip[maxlag * stepout:100 -
(maxlag * stepout), stepout * maxlag:100 -
(stepout * maxlag)] +
test_data[maxlag * stepout:100 -
(maxlag * stepout), maxlag * stepout:100 -
(stepout * maxlag)])
diff = diff[1:-1, 1:-1]
# Output should be all -1 except for the edge effects
same = np.allclose(diff, np.zeros(diff.shape), .01)
self.assertTrue(same)
def test_wavelet(self):
# Make a dipping reflectivity field
data = np.zeros( ( 150,100 ) )
data[0,:] = np.arange( 100 ) / 10.0
dips = np.arange( 0,100,2)
dips = np.concatenate( [dips, dips[::-1] ] )
window_size = 10
freq = 20.0
duration = .2
wavelet = ricker( window_size, 1, freq )
for i in range( dips.size ):
data[:,i]= np.convolve( wavelet, np.roll( data[:,i],
dips[i] ),
mode = 'same' )
stepout=1
maxlag=2
dip = dipsteer( data, window_size, stepout=stepout,
maxlag=maxlag, overlap=0.5 )
undipped = np.zeros( dips.shape )
for i in range ( dips.size ):
undipped[i] = ( dip[dips[i],i] )
check = undipped[ maxlag*stepout:-maxlag*stepout ]
test = np.zeros( check.shape )
test[0 : test.size/2 -1] = -2.0
test[ (test.size/2) - 1 ] = -1
test[(test.size/2) ] = 1
test[ (test.size/2)+1 : ] = 2
"""fig = pl.figure()
pl.imshow( dip )
fig.show()
fig2 = pl.figure()
pl.imshow( data)
fig2.show()"""
self.assertTrue( np.allclose( test, check ) )