Convert imaging data



In [1]:

    
import matplotlib.pyplot as plt
%matplotlib inline



In [2]:

    
from showit import image



In [3]:

    
from skimage.io import imsave



In [4]:

    
from os.path import exists
from os import mkdir
from json import dump

Set directory and session information



In [5]:

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

Load data



In [7]:

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









    



/tier2/freeman/Nick/sessions/000222
True



In [8]:

    
import mesoscope as ms



In [9]:

    
import thunder as td



In [10]:

    
rawData = td.images.fromtif(path + '/images-raw/', nplanes=1, engine = sc)



In [81]:

    
foo = rawData.values.reshape((8753,8,300,300), 'C')



In [42]:

    
bar = td.images.Images(foo)



In [44]:

    
bar.count()









    Out[44]:





70024



In [28]:

    
rawDataSm = rawData[15:183]



In [33]:

    
data = rawDataSm



In [30]:

    
t, z = rawDataSm.shape[0]/8, 8
x, y = rawData.shape[1:]
#swapped = rawData.values.reshape(z, t, x, y).transpose(1, 0, 2, 3).swap((1), ())
swapped = rawDataSm.values.reshape(t, z, x, y).swap((1), ())
reshaped = td.images.Images(swapped)



In [31]:

    
img = reshaped.mean().toarray()



In [11]:

    
rawData, rawMetadata = ms.load(path + '/images-raw/', engine = sc)
data, meta = ms.convert(rawData, rawMetadata)



In [ ]:



In [13]:

    
print data
print '\n', meta









    



Images
mode: spark
dtype: int16
shape: (202, 2100, 2100)

{'nrois': 1, 'power': 100, 'nchannels': 1, 'nplanes': 1, 'averaging': 1000, 'depths': 200, 'rois': [{'depths': 0.0, 'npixels': [2100.0, 2100.0], 'center': [-3.1720657846433042e-17, 0.0], 'size': [4.2000000005249998, 4.2000000005250007]}], 'shape': (202, 2100, 2100), 'volumeRate': 1.59768655}

Inspect data



In [90]:

    
#img = data.mean().toarray()
#img = data[0].toarray()



In [26]:

    
image(img[2], clim=(0, 3.5*img.mean()), size = 10);

Save binary images



In [34]:

    
if not exists(path + '/images'):
    mkdir(path + '/images')
data.tobinary(path + '/images', overwrite=True)

Save summary data

Meta data



In [31]:

    
if not exists(path + '/summary'):
    mkdir(path + '/summary')
with open(path + '/summary/meta.json', 'w') as fid:
    dump(meta, fid, indent=2)









    



---------------------------------------------------------------------------
OSError                                   Traceback (most recent call last)
<ipython-input-31-241effffe23f> in <module>()
      1 if not exists(path + '/summary'):
----> 2     mkdir(path + '/summary')
      3 #with open(path + '/summary/meta.json', 'w') as fid:
      4 #    dump(meta, fid, indent=2)

OSError: [Errno 17] File exists: '/tier2/freeman/Nick/sessions/000207/summary'

Raw tiffs



In [32]:

    
imsave(path + '/summary/rawVol.tif', (img.clip(0,2**13-1)/32).astype('uint8'), plugin='tifffile', photometric='minisblack')

Normalized tiffs



In [18]:

    
from numpy import array
def normalize(oim):
    """
    Normalize a three dimensional volume.
    """
    if not oim.ndim == 3:
        raise Exception('image must be three dimensional')
    means = oim.mean(axis=(1, 2), dtype='float32')
    maximum = means.max()
    return array([oim[i]*maximum/means[i] for i in range(oim.shape[0])])



In [19]:

    
norm = normalize(img).astype('float32')
#norm = data.map(normalize)



In [21]:

    
imsave(path + '/summary/raw-norm.tif', (norm.clip(0,2**13-1)/32).astype('uint8'), plugin='tifffile', photometric='minisblack')



In [ ]: