Delete a lot of scripts made just for plotting. Create draft modules …

…in which part of the functionality of these scripts will go.

derotation/postprocessing/contrast_enhancement.py

@@ -0,0 +1,35 @@
+from pathlib import Path
+
+import numpy as np
+import tifffile as tf
+from skimage.exposure import rescale_intensity
+
+
+def whole_video_contrast_enhancment(
+    tiff_path: Path, saturated_percentage: float
+):
+    tiff = tf.imread(tiff_path)
+    v_min, v_max = np.percentile(
+        tiff, (saturated_percentage, 100 - saturated_percentage)
+    )
+
+    CE_tiff_path = np.zeros_like(tiff)
+    for i, frame in enumerate(tiff):
+        CE_tiff_path[i] = rescale_intensity(frame, in_range=(v_min, v_max))
+
+    return CE_tiff_path
+
+
+def frame_by_frame_contrast_enhancement(
+    tiff_path: Path, saturated_percentage: float
+):
+    tiff = tf.imread(tiff_path)
+
+    CE_tiff_path = np.zeros_like(tiff)
+    for i, frame in enumerate(tiff):
+        v_min, v_max = np.percentile(
+            frame, (saturated_percentage, 100 - saturated_percentage)
+        )
+        CE_tiff_path[i] = rescale_intensity(frame, in_range=(v_min, v_max))
+
+    return CE_tiff_path

derotation/postprocessing/dashboard.py

@@ -0,0 +1,235 @@
+import warnings
+from pathlib import Path
+
+import allensdk.brain_observatory.dff as dff_module
+import fastplotlib as fpl
+import numpy as np
+import pandas as pd
+import pynapple as nap
+from allensdk.brain_observatory.r_neuropil import NeuropilSubtract
+from ipywidgets import AppLayout
+from skimage.measure import find_contours
+
+warnings.filterwarnings("ignore", category=DeprecationWarning)
+
+# load data
+
+suite2p_data = nap.load_file("suite2p.nwb")
+rotation_info = pd.read_csv("rotation_info.csv")
+
+
+# Isolate first main objects
+# fluorescence
+raw_fluorescence = suite2p_data["RoiResponseSeries"]
+
+# neuropil
+neuropil = suite2p_data["Neuropil"]
+
+# rotation_angles
+rotation_angles = nap.Tsd(
+    t=raw_fluorescence.t, d=rotation_info["rotation_angle"]
+)
+
+# when is the rotation happening: interval set
+rotation_stim = nap.TsdFrame(
+    t=raw_fluorescence.t,
+    d=rotation_info[["direction", "speed", "rotation_count"]].values,
+    columns=["direction", "speed", "rotation_count"],
+)
+
+rotation_on_intervals = rotation_angles.threshold(0).time_support.union(
+    rotation_angles.threshold(0, "below").time_support
+)
+
+stimuli_start_info = nap.TsdFrame(
+    rotation_stim.restrict(rotation_on_intervals)
+    .as_dataframe()
+    .drop_duplicates("rotation_count")
+    .dropna()
+)
+
+
+def neuropil_subtraction(f, f_neu):
+    #  use default parameters for all methods
+    neuropil_subtraction = NeuropilSubtract()
+    neuropil_subtraction.set_F(f, f_neu)
+    neuropil_subtraction.fit()
+
+    r = neuropil_subtraction.r
+
+    f_corr = f - r * f_neu
+
+    dff = 100 * dff_module.compute_dff_windowed_median(
+        f_corr, median_kernel_long=1213, median_kernel_short=23
+    )
+
+    return dff, r
+
+
+dff, r = neuropil_subtraction(
+    raw_fluorescence[:].values.T, neuropil[:].values.T
+)
+
+dff = nap.dff(t=raw_fluorescence.t, d=dff.T)
+
+stimuli_start_info_df = stimuli_start_info.as_dataframe()
+rotation_on_intervals_df = rotation_on_intervals.as_dataframe()
+for key in stimuli_start_info_df.keys()[:2]:
+    for item in stimuli_start_info_df[key].unique():
+        subset = stimuli_start_info_df[stimuli_start_info_df[key] == item]
+        starts = subset.index
+        rotation_on_intervals_df[rotation_on_intervals_df["start"] == starts]
+
+# where is the data in nwb file?
+plane_seg = (
+    suite2p_data.nwb.processing["ophys"]
+    .data_interfaces["ImageSegmentation"]
+    .plane_segmentations["PlaneSegmentation"]
+)
+
+ROI_centroids = plane_seg.ROICentroids[:]
+is_cell = plane_seg.Accepted[:]
+labels = plane_seg.image_mask[:]
+
+
+contours = [find_contours(c)[0] for c in labels[is_cell.astype(bool)]]
+
+
+# import the registered binary
+path_to_bin_file = Path(
+    "/Users/laura/data/derotation/raw/230802_CAA_1120182/derotated/archive/test/suite2p/plane0/data.bin"
+)
+
+shape_image = (16516, 256, 256)
+registered = np.memmap(path_to_bin_file, shape=shape_image, dtype="int16")
+
+
+def find_ROI(nearest):
+    x, y, _ = np.asarray(nearest.data.value).mean(axis=0)
+    # print(f'x:{x}, y:{y}')
+    closest_x = np.abs(ROI_centroids[:, 0] - x).argmin()
+    closest_y = np.abs(ROI_centroids[:, 1] - y).argmin()
+    # print(f'closest x: {closest_x}, closest_y: {closest_y}')
+    if closest_x != closest_y:
+        delta_x = ROI_centroids[closest_x, 0] - x
+        delta_y = ROI_centroids[closest_y, 1] - y
+        if delta_x < delta_y:
+            return closest_x
+        else:
+            return closest_y
+    else:
+        return closest_x
+
+
+# Panel 1
+colors = {key: "white" for key in range(len(rotation_info["rotation_angle"]))}
+for i, row in rotation_info.iterrows():
+    if row["direction"] == -1:
+        colors[i] = "green"
+    if row["direction"] == 1:
+        colors[i] = "magenta"
+
+rotation_fig = fpl.Figure((3, 1), size=(1200, 300))
+angles_top_plot = rotation_fig[0, 0].add_line(
+    data=rotation_info["rotation_angle"],
+    thickness=1,
+    colors=list(colors.values()),
+)
+
+region_selector = angles_top_plot.add_linear_region_selector()
+zoomed_init = region_selector.get_selected_data()
+zoomed_x = rotation_fig[1, 0].add_line(zoomed_init)
+
+selector = angles_top_plot.add_linear_selector()
+
+selection_boundaries = region_selector.get_selected_indices()
+sliced_dff = dff[selection_boundaries[0] : selection_boundaries[-1]]
+
+dff_lines = rotation_fig[2, 0].add_line_stack(
+    np.asarray(sliced_dff).T[is_cell],
+    cmap="jet",
+    thickness=1,
+    separation=1,
+)
+
+
+@region_selector.add_event_handler("selection")
+def slice_dff_array(ev):
+    global zoomed_x
+    selected_data = ev.get_selected_data()
+    rotation_fig[1, 0].remove_graphic(zoomed_x)
+    zoomed_x = rotation_fig[1, 0].add_line(selected_data)
+    rotation_fig[1, 0].auto_scale()
+
+    global dff_lines
+    global sliced_dff
+    selection_boundaries = ev.get_selected_indices()
+    sliced_dff = dff[selection_boundaries[0] : selection_boundaries[-1]]
+
+    rotation_fig[2, 0].clear()
+    dff_lines = rotation_fig[2, 0].add_line_stack(
+        np.asarray(sliced_dff).T[is_cell],
+        cmap="jet",
+        thickness=1,
+        separation=1,
+    )
+    rotation_fig[2, 0].auto_scale()
+
+    fig_single_trace[0, 0].clear()
+    fig_single_trace[0, 0].add_line(
+        data=np.asarray(sliced_dff).T[idx],
+        thickness=2,
+        cmap="plasma",
+    )
+    fig_single_trace[0, 0].auto_scale()
+
+
+# Panel 2
+iw = fpl.ImageWidget(
+    registered[: len(rotation_info["rotation_angle"])], cmap="gnuplot2"
+)
+iw.show()
+contours_graphic = iw.figure[0, 0].add_line_collection(
+    contours, thickness=2, colors="green"
+)
+
+# Panel 3
+fig_single_trace = fpl.Figure()
+idx = 0
+single_line = fig_single_trace[0, 0].add_line(
+    data=np.asarray(sliced_dff).T[idx],
+    thickness=2,
+    cmap="plasma",
+)
+
+rotation_fig.show(maintain_aspect=False)
+
+
+def set_selected_component(ev):
+    xy = iw.figure[0, 0].map_screen_to_world(ev)[:-1]
+    global nearest
+    nearest = fpl.utils.get_nearest_graphics(xy, contours_graphic)[0]
+    contours_graphic.colors = "green"
+    nearest.colors = "w"
+
+    global idx
+    global single_line
+    idx = find_ROI(nearest)
+    fig_single_trace[0, 0].clear()
+    single_line = fig_single_trace[0, 0].add_line(
+        data=np.asarray(sliced_dff).T[idx],
+        thickness=2,
+        cmap="plasma",
+    )
+    fig_single_trace[0, 0].auto_scale()
+
+
+iw.figure.renderer.add_event_handler(set_selected_component, "click")
+selector.add_ipywidget_handler(iw.sliders["t"], step=1)
+
+# Handling ipywidgets layout
+AppLayout(
+    header=rotation_fig.show(maintain_aspect=False),
+    left_sidebar=iw.show(),
+    right_sidebar=fig_single_trace.show(maintain_aspect=False),
+)

derotation/postprocessing/neuropil_subtraction.py

@@ -0,0 +1,21 @@
+import allensdk.brain_observatory.dff as dff_module
+from allensdk.brain_observatory.r_neuropil import NeuropilSubtract
+
+
+def neuropil_subtraction(f, f_neu):
+    #  use default parameters for all methods
+    neuropil_subtraction = NeuropilSubtract()
+    neuropil_subtraction.set_F(f, f_neu)
+    neuropil_subtraction.fit()
+
+    r = neuropil_subtraction.r
+
+    f_corr = f - r * f_neu
+
+    # kernel values to be changed for 3-photon data
+
+    dff = 100 * dff_module.compute_dff_windowed_median(
+        f_corr, median_kernel_long=1213, median_kernel_short=23
+    )
+
+    return dff, r

derotation/postprocessing/registration.py

@@ -0,0 +1,2 @@
+def get_registration_metrics():
+    pass