Skip to content

Commit

Permalink
add tests for stack.preprocessing and stack.utils
Browse files Browse the repository at this point in the history
  • Loading branch information
@Henley13
Henley13 committed May 5, 2020
1 parent 8fc4b38 commit a4a36de
Show file tree
Hide file tree
Showing 5 changed files with 346 additions and 124 deletions.
8 changes: 4 additions & 4 deletions bigfish/stack/__init__.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,8 +5,8 @@
build stack of images.
"""

from .utils import (check_array, check_df, check_recipe, check_parameter,
check_range_value,
from .utils import (check_array, check_df, check_recipe, check_datamap,
check_parameter, check_range_value,
get_margin_value, get_eps_float32)
from .io import (read_image, read_dv, read_array, save_image, save_array)
from .preprocess import (build_stacks, build_stack,
Expand Down Expand Up @@ -38,8 +38,8 @@
from .augmentation import augment


_utils = ["check_array", "check_df", "check_recipe", "check_parameter",
"check_range_value",
_utils = ["check_array", "check_df", "check_recipe", "check_datamap",
"check_parameter", "check_range_value",
"get_margin_value", "get_eps_float32"]

_io = ["read_image", "read_dv", "read_array", "save_image", "save_array"]
Expand Down
140 changes: 55 additions & 85 deletions bigfish/stack/preprocess.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@
import numpy as np

from .io import read_image
from .utils import (check_array, check_parameter, check_recipe,
from .utils import (check_array, check_parameter, check_recipe, check_datamap,
check_range_value, fit_recipe,
get_path_from_recipe, get_nb_element_per_dimension,
count_nb_fov)
Expand All @@ -20,8 +20,6 @@
from skimage.exposure import rescale_intensity


# TODO be able to build only one channel

# ### Building stack ###

def build_stacks(data_map, input_dimension=None, sanity_check=False,
Expand All @@ -47,17 +45,17 @@ def build_stacks(data_map, input_dimension=None, sanity_check=False,
The recipe dictionary for one field of view takes the form:
{
"fov": List[str], (optional)
"z": List[str], (optional)
"c": List[str], (optional)
"r": List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern"
"fov": str or List[str], (optional)
"z": str or List[str], (optional)
"c": str or List[str], (optional)
"r": str or List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern": str
}
- A field of view is defined by an ID common to every images belonging to
the same field of view ("fov").
- A field of view is defined by a string common to every images belonging
to the same field of view ("fov").
- At least every images are in 2-d with x and y dimensions. So we need to
mention the round-dimension, the channel-dimension and the z-dimension to
add ("r", "c" and "z"). For these keys, we provide a list of
Expand All @@ -66,7 +64,6 @@ def build_stacks(data_map, input_dimension=None, sanity_check=False,
can be provided with the file extension "ext" (usually 'tif' or 'tiff') or
an optional morpheme ("opt").
- A pattern used to get the filename ("pattern").
- The fields "fov", "z", "c" and "r" can be strings instead of lists.
Example 1. Let us assume 3-d images (zyx dimensions) saved as
"r03c03f01_405.tif", "r03c03f01_488.tif" and "r03c03f01_561.tif". The first
Expand Down Expand Up @@ -116,46 +113,51 @@ def build_stacks(data_map, input_dimension=None, sanity_check=False,
Path of the input directory from where the tensor is built.
recipe : dict
Recipe used to build the tensor.
i_fov : int
Index of the fov to build (for a specific recipe).
"""
# check parameters
check_parameter(data_map=list,
input_dimension=(int, type(None)),
sanity_check=bool,
return_origin=bool)
check_datamap(data_map)

# load and generate tensors for each recipe-folder pair
for recipe, input_folder in data_map:

# load and generate tensors for each fov stored in a recipe
nb_fov = count_nb_fov(recipe)
for i_fov in range(nb_fov):
tensor = build_stack(recipe, input_folder, input_dimension, i_fov,
sanity_check)
tensor = build_stack(recipe, input_folder,
input_dimension=input_dimension,
sanity_check=sanity_check,
i_fov=i_fov)
if return_origin:
yield tensor, input_folder, recipe, i_fov
else:
yield tensor


def build_stack(recipe, input_folder, input_dimension=None, i_fov=0,
sanity_check=False):
"""Build 5-d stack and normalize it.
def build_stack(recipe, input_folder, input_dimension=None, sanity_check=False,
i_fov=0):
"""Build a 5-d stack from the same field of view (fov).
The recipe dictionary for one field of view takes the form:
{
"fov": List[str], (optional)
"z": List[str], (optional)
"c": List[str], (optional)
"r": List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern"
"fov": str or List[str], (optional)
"z": str or List[str], (optional)
"c": str or List[str], (optional)
"r": str or List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern": str
}
- A field of view is defined by an ID common to every images belonging to
the same field of view ("fov").
- A field of view is defined by a string common to every images belonging
to the same field of view ("fov").
- At least every images are in 2-d with x and y dimensions. So we need to
mention the round-dimension, the channel-dimension and the z-dimension to
add ("r", "c" and "z"). For these keys, we provide a list of
Expand All @@ -164,7 +166,6 @@ def build_stack(recipe, input_folder, input_dimension=None, i_fov=0,
can be provided with the file extension "ext" (usually 'tif' or 'tiff') or
an optional morpheme ("opt").
- A pattern used to get the filename ("pattern").
- The fields "fov", "z", "c" and "r" can be strings instead of lists.
Example 1. Let us assume 3-d images (zyx dimensions) saved as
"r03c03f01_405.tif", "r03c03f01_488.tif" and "r03c03f01_561.tif". The first
Expand Down Expand Up @@ -240,31 +241,30 @@ def build_stack(recipe, input_folder, input_dimension=None, i_fov=0,


def _load_stack(recipe, input_folder, input_dimension=None, i_fov=0):
"""Build a 5-d tensor from the same fields of view (fov).
"""Build a 5-d tensor from the same field of view (fov).
The function stacks a set of images using a recipe mapping the
different images with the dimensions they represent. Each stacking step
add a new dimension to the original tensors (eg. we stack 2-d images with
the same xy coordinates, but different depths to get a 3-d image). If the
files we need to build a new dimension are not included in the
recipe, an empty dimension is added. This operation is repeated until we
get a 5-d tensor. We first operate on the z dimension, then the
channels and eventually the rounds.
the same xy coordinates to get a 3-d image). If the files we need to build
a new dimension are not included in the recipe, an empty dimension is
added. This operation is repeated until we get a 5-d tensor. We first
operate on the z dimension, then the channels and eventually the rounds.
The recipe dictionary for one field of view takes the form:
{
"fov": List[str], (optional)
"z": List[str], (optional)
"c": List[str], (optional)
"r": List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern"
"fov": str or List[str], (optional)
"z": str or List[str], (optional)
"c": str or List[str], (optional)
"r": str or List[str], (optional)
"ext": str, (optional)
"opt": str, (optional)
"pattern": str
}
- A field of view is defined by an ID common to every images belonging to
the same field of view ("fov").
- A field of view is defined by a string common to every images belonging
to the same field of view ("fov").
- At least every images are in 2-d with x and y dimensions. So we need to
mention the round-dimension, the channel-dimension and the z-dimension to
add ("r", "c" and "z"). For these keys, we provide a list of
Expand All @@ -273,36 +273,6 @@ def _load_stack(recipe, input_folder, input_dimension=None, i_fov=0):
can be provided with the file extension "ext" (usually 'tif' or 'tiff') or
an optional morpheme ("opt").
- A pattern used to get the filename ("pattern").
- The fields "fov", "z", "c" and "r" can be strings instead of lists.
Example 1. Let us assume 3-d images (zyx dimensions) saved as
"r03c03f01_405.tif", "r03c03f01_488.tif" and "r03c03f01_561.tif". The first
morpheme "r03c03f01" uniquely identifies a 3-d field of view. The second
morphemes "405", "488" and "561" identify three different channels we
want to stack. There is no round in this experiment. We need to return a
tensor with shape (1, 3, z, y, x). Thus, a valid recipe would be:
{
"fov": "r03c03f01",
"c": ["405", "488", "561"],
"ext": "tif"
"pattern": "fov_c.ext"
}
Example 2. Let us assume 2-d images (yx dimensions) saved as
"dapi_1.TIFF", "cy3_1.TIFF", "GFP_1.TIFF", "dapi_2.TIFF", "cy3_2.TIFF" and
"GFP_2.TIFF". The first morphemes "dapi", "cy3" and "GFP" identify
channels. The second morphemes "1" and "2" identify two different fields of
view. There is no round and no z dimension in this experiment. We can
build two tensors with shape (1, 3, 1, y, x). Thus, a valid recipe would
be:
{
"fov": ["1", "2"],
"c": ["dapi", "cy3", "GFP"],
"ext": "TIFF"
"pattern": "c_fov.ext"
}
Parameters
----------
Expand All @@ -319,8 +289,8 @@ def _load_stack(recipe, input_folder, input_dimension=None, i_fov=0):
Returns
-------
stack : np.ndarray, np.uint
Tensor with shape (r, c, z, y, x).
stack : np.ndarray
Tensor with shape (round, channel, z, y, x).
"""
# complete the recipe with unused morphemes
Expand All @@ -347,8 +317,8 @@ def _load_stack(recipe, input_folder, input_dimension=None, i_fov=0):
stack = _build_stack_from_5d(recipe, input_folder, fov=i_fov)
else:
raise ValueError("Files do not have the right number of dimensions: "
"{0}. The files we stack should be in 2-d, 3-d, 4-d "
"or 5-d.".format(input_dimension))
"{0}. The files we stack should have between 2 and "
"5 dimensions.".format(input_dimension))

return stack

Expand All @@ -375,8 +345,8 @@ def _build_stack_from_2d(recipe, input_folder, fov=0, nb_r=1, nb_c=1, nb_z=1):
Returns
-------
tensor_5d : np.ndarray, np.uint
Tensor with shape (r, c, z, y, x).
tensor_5d : np.ndarray
Tensor with shape (round, channel, z, y, x).
"""

Expand Down Expand Up @@ -430,8 +400,8 @@ def _build_stack_from_3d(recipe, input_folder, fov=0, nb_r=1, nb_c=1):
Returns
-------
tensor_5d : np.ndarray, np.uint
Tensor with shape (r, c, z, y, x).
tensor_5d : np.ndarray
Tensor with shape (round, channel, z, y, x).
"""
# load and stack successively channel elements then round elements
Expand Down Expand Up @@ -474,8 +444,8 @@ def _build_stack_from_4d(recipe, input_folder, fov=0, nb_r=1):
Returns
-------
tensor_5d : np.ndarray, np.uint
Tensor with shape (r, c, z, y, x).
tensor_5d : np.ndarray
Tensor with shape (round, channel, z, y, x).
"""
# load each file from a new round element and stack them
Expand Down Expand Up @@ -507,8 +477,8 @@ def _build_stack_from_5d(recipe, input_folder, fov=0):
Returns
-------
tensor_5d : np.ndarray, np.uint
Tensor with shape (r, c, z, y, x).
tensor_5d : np.ndarray
Tensor with shape (round, channel, z, y, x).
"""
# the recipe can only contain one file with a 5-d tensor per fov
Expand Down
Loading

0 comments on commit a4a36de

Please sign in to comment.