def convert_to_rgb(df, split, resize=True, new_size=224, extension='jpeg'):
N = df.shape[0]
for i in tqdm(range(N)):
code = df['id_code'][i]
experiment = df['experiment'][i]
plate = df['plate'][i]
well = df['well'][i]
for site in [1, 2]:
save_path = os.path.join(
BASE_DIR, '{}/{}_s{}.{}'.format(split, code, site, extension))
im = rio.load_site_as_rgb(split,
experiment,
plate,
well,
site,
base_path=ORI_DIR)
im = im.astype(np.uint8)
im = Image.fromarray(im)
if resize:
im = im.resize((new_size, new_size), resample=Image.BILINEAR)
im.save(save_path)
def moving_files(df=x_train_df, mode='train'):
indexes = df.index
bar = Bar(mode + '_processing', max=len(df))
for i in tqdm.tqdm(range(len(df))):
row = df.iloc[i, :]
class_name = str(int(row['sirna']))
create_folder(mode + '_data/' + class_name)
img = rio.load_site_as_rgb('train', row['experiment'], row['plate'],
row['well'], row['site'])
img = img.astype(np.uint8)
dest = mode + '_data/' + class_name + '/' + indexes[i] + '_' + str(
row['site']) + '.png'
plt.imsave(dest, img)
bar.next()
bar.finish()
def moving_files(df,t,mode='test'):
indexes = df.index
bar = Bar(t, max=len(df))
for i in tqdm.tqdm(range(len(df))):
row = df.iloc[i,:]
img = rio.load_site_as_rgb(mode,
row['experiment'],
row['plate'],
row['well'],
row['site'])
img = img.astype(np.uint8)
dest = mode+'_data/test/'+indexes[i]+'_'+str(row['site'])+'.png'
bar.next()
plt.imsave(dest,img)
bar.finish()
def convert_to_rgb(df, split, resize=True, new_size=400, extension='jpeg'):
N = df.shape[0]
for i in tqdm(range(N)):
code = df['id_code'][i]
experiment = df['experiment'][i]
plate = df['plate'][i]
well = df['well'][i]
for site in [1, 2]:
save_path = f'rxrxrgb/{split}/{code}_s{site}.{extension}'
im = rio.load_site_as_rgb(
split, experiment, plate, well, site,
base_path=data_path
)
im = im.astype(np.uint8)
im = Image.fromarray(im)
if resize:
im = im.resize((new_size, new_size), resample=Image.BILINEAR)
im.save(save_path)
def convert_to_rgb_(cell):
split = cell.split
code = cell.code
site = cell.site
extension = cell.extension
experiment = cell.experiment
plate = cell.plate
well = cell.well
save_path = f'{split}/{code}_s{site}.{extension}'
im = rio.load_site_as_rgb(split,
experiment,
plate,
well,
site,
base_path='./data/')
im = im.astype(np.uint8)
im = Image.fromarray(im)
# im = im.resize((512, 512), resample=Image.BILINEAR)
im.save(save_path)
def convert_to_rgb(df, split, resize=False, new_size=224, extension='png'):
RGB_MAP = {
1: {
'rgb': np.array([19, 0, 249]),
'range': [0, 51]
},
2: {
'rgb': np.array([42, 255, 31]),
'range': [0, 107]
},
3: {
'rgb': np.array([255, 0, 25]),
'range': [0, 64]
},
4: {
'rgb': np.array([45, 255, 252]),
'range': [0, 191]
},
5: {
'rgb': np.array([250, 0, 253]),
'range': [0, 89]
},
6: {
'rgb': np.array([254, 255, 40]),
'range': [0, 191]
}
}
N = df.shape[0]
print(df.head())
print(df['id_code'].iloc[1])
for i in tqdm(range(N)):
code = df['id_code'].iloc[i]
experiment = df['experiment'].iloc[i]
plate = df['plate'].iloc[i]
well = df['well'].iloc[i]
sirna = df['sirna'].iloc[i]
for site in [1, 2]:
folder = '../input/' + str(split) + '/' + str(sirna) + '/'
if not os.path.exists(folder):
os.makedirs(folder)
save_path = '../input/' + str(split) + '/' + str(
sirna) + '/' + str(code) + '_s' + str(site) + '.' + str(
extension)
im = rio.load_site_as_rgb(split,
experiment,
plate,
well,
site,
channels=(1, 2, 3, 4, 5, 6),
base_path='../input/',
rgb_map=RGB_MAP)
im = im.astype(np.uint8)
im = Image.fromarray(im)
if resize:
im = im.resize((new_size, new_size), resample=Image.BILINEAR)
im.save(save_path)
x = rio.convert_tensor_to_rgb(t)
x.shape
# In[7]:
plt.figure(figsize=(8, 8))
plt.axis('off')
_ = plt.imshow(x)
# In[8]:
y = rio.load_site_as_rgb('train',
'HUVEC-08',
4,
'K09',
1,
base_path=LOCAL_IMAGES_BASE_PATH)
plt.figure(figsize=(8, 8))
plt.axis('off')
_ = plt.imshow(y)
# In[9]:
# train.csv, train_controls.csv, test.csv, test_controls.csv
md = rio.combine_metadata(base_path=DEFAULT_METADATA_BASE_PATH)
md.head()
import rxrx.io as rio
t = rio.load_site('train', 'RPE-05', 3, 'D19', 2)
t.shape
fig, axes = plt.subplots(2, 3, figsize=(24, 16))
for i, ax in enumerate(axes.flatten()):
ax.axis('off')
ax.set_title('channel {}'.format(i + 1))
_ = ax.imshow(t[:, :, i], cmap='gray')
x = rio.convert_tensor_to_rgb(t)
x.shape
plt.figure(figsize=(8, 8))
plt.axis('off')
_ = plt.imshow(x)
y = rio.load_site_as_rgb('train', 'HUVEC-08', 4, 'K09', 1)
plt.figure(figsize=(8, 8))
plt.axis('off')
_ = plt.imshow(y)
md = rio.combine_metadata()
md.head()
import seaborn as sns
md.head(10)
md.index
for i in md.columns:
print(">> ", i, "\t", md[i].unique())
for col in [
'cell_type', 'dataset', 'experiment', 'plate', 'site', 'well_type'
]:
