Setup



In [2]:

    
%matplotlib inline



In [3]:

    
from numpy import array
import matplotlib.pyplot as plt
import seaborn as sns



In [4]:

    
from showit import image, tile
import matplotlib.animation as animation



In [5]:

    
from os.path import join, exists
from os import mkdir, makedirs



In [6]:

    
#from mesoscope.utils import normalize



In [7]:

    
import thunder



In [8]:

    
from registration import CrossCorr

Set directory and session information



In [9]:

    
directory = '/tier2/freeman/Nick/sessions'



In [10]:

    
key = '000222'
prefix = 'trial'
path = directory + '/' + key
print path
print exists(path)









    



/tier2/freeman/Nick/sessions/000222
True

Load data



In [11]:

    
data = thunder.images.fromtif(path + '/images', engine=None)



In [12]:

    
data









    Out[12]:





Images
mode: local
dtype: uint8
shape: (202, 2100, 2100)

Register data



In [1]:

    
from mesoscope import reference



In [13]:

    
ref = reference(data)



In [14]:

    
fig = plt.figure(figsize=[10,10])
ax = plt.axes()
im = image(ref, clim=(0, 3.5*ref.mean()), ax=ax)



In [15]:

    
ref.shape









    Out[15]:





(2100, 2100)



In [ ]:



In [16]:

    
algorithm = CrossCorr(axis=0)



In [17]:

    
model = algorithm.fit(data, reference)



In [18]:

    
registered = model.transform(data)



In [19]:

    
registered









    Out[19]:





Images
mode: spark
dtype: int16
shape: (8585, 4, 512, 512)

Inspect shifts



In [20]:

    
shifts = model.toarray()



In [21]:

    
shifts.shape









    Out[21]:





(8585, 4, 2)



In [22]:

    
[plt.plot(model.toarray()[:,i]) for i in range(shifts.shape[1])];
plt.xlim([0, 1000])









    Out[22]:





(0, 1000)

Inspect data



In [23]:

    
mean = registered.mean().toarray()



In [24]:

    
mean.shape









    Out[24]:





(4, 512, 512)



In [25]:

    
img = mean[0]
fig = plt.figure(figsize=[10,10])
ax = plt.axes()
im = image(img, clim=(0, 3.5*img.mean()), ax=ax)

Save registered binary data



In [26]:

    
if not exists(path + '/registered'):
    makedirs(path + '/registered')
registered.tobinary(path + '/registered', overwrite=True)

Save summary data



In [27]:

    
if not exists(path + '/summary'):
    makedirs(path + '/summary')

Raw tiffs



In [28]:

    
from skimage.io import imsave
imsave(path + '/summary/registered.tif', mean.astype('float32'), plugin='tifffile', photometric='minisblack')
imsave(path + '/summary/reference.tif', reference.astype('float32'), plugin='tifffile', photometric='minisblack')

Normalized tiffs



In [29]:

    
imsave(path + '/summary/registered-norm.tif', normalize(mean).astype('float32'), plugin='tifffile', photometric='minisblack')
imsave(path + '/summary/reference-norm.tif', normalize(reference).astype('float32'), plugin='tifffile', photometric='minisblack')