Skip to content

Commit

Permalink
napari analysis docs
Browse files Browse the repository at this point in the history
  • Loading branch information
@aelefebv
aelefebv committed Oct 3, 2024
1 parent 310bc06 commit cc8b206
Showing 1 changed file with 194 additions and 0 deletions.
194 changes: 194 additions & 0 deletions nellie_napari/nellie_analysis.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,98 @@


class NellieAnalysis(QWidget):
"""
A class for analyzing and visualizing multi-dimensional microscopy data using histograms, graphs, and overlays in the napari viewer.
Attributes
----------
viewer : napari.viewer.Viewer
An instance of the napari viewer.
nellie : object
Reference to the main Nellie object that contains the pipeline and analysis data.
canvas : FigureCanvasQTAgg
Canvas for rendering matplotlib plots.
scale : tuple
The scaling factors for X, Y, and Z dimensions, default is (1, 1, 1).
log_scale : bool
Boolean flag to toggle logarithmic scaling in histogram plots.
is_median : bool
Boolean flag to toggle between mean and median views.
match_t : bool
Boolean flag to enable/disable timepoint matching in data analysis.
hist_reset : bool
Boolean flag indicating whether the histogram settings are reset.
voxel_df, node_df, branch_df, organelle_df, image_df : pd.DataFrame
DataFrames containing features at different hierarchy levels (voxel, node, branch, organelle, image).
attr_data : pd.Series or None
Data for the selected attribute to be plotted.
time_col : pd.Series or None
Time column data from the currently selected level's DataFrame.
adjacency_maps : dict or None
Dictionary containing adjacency matrices for mapping hierarchy levels.
mean, std, median, iqr, perc75, perc25 : float
Statistical values for the selected attribute data (mean, std, median, interquartile range, 75th percentile, 25th percentile).
Methods
-------
reset()
Resets the internal state, including dataframes and initialization flags.
post_init()
Initializes UI elements such as checkboxes, buttons, and connects them to respective event handlers.
set_ui()
Sets up the user interface layout, including attribute selection, histogram, and export options.
_create_dropdown_selection()
Creates and configures the dropdowns for selecting hierarchy levels and attributes.
set_default_dropdowns()
Sets default values for the hierarchy level and attribute dropdowns.
check_for_adjacency_map()
Checks if an adjacency map is available and enables the overlay button accordingly.
rewrite_dropdown()
Rewrites the dropdown options based on the available data and features.
export_data()
Exports the current graph data to a CSV file.
save_graph()
Saves the current graph as a PNG file.
on_hist_change(event)
Event handler for histogram changes (e.g., adjusting bins or min/max values).
get_index(layer, event)
Gets the voxel index based on mouse hover coordinates in the napari viewer.
overlay()
Applies an overlay of selected attribute data onto the image, using adjacency maps to map voxel to higher-level features.
on_t_change(event)
Event handler that updates the graph when the timepoint changes in the napari viewer.
toggle_match_t(state)
Toggles timepoint matching and updates the graph accordingly.
toggle_mean_med(state)
Toggles between mean and median views and updates the graph.
get_csvs()
Loads the feature CSV files into DataFrames for voxels, nodes, branches, organelles, and images.
on_level_selected(index)
Event handler for when a hierarchy level is selected from the dropdown.
on_attr_selected(index)
Event handler for when an attribute is selected from the dropdown.
get_stats()
Computes basic statistics (mean, std, median, percentiles) for the selected attribute data.
draw_stats()
Draws the computed statistics on the histogram plot (e.g., mean, std, median, percentiles).
plot_data(title)
Plots the selected attribute data as a histogram, updates the UI, and displays statistical information.
on_log_scale(state)
Toggles logarithmic scaling for the histogram plot and refreshes the data.
"""
def __init__(self, napari_viewer: 'napari.viewer.Viewer', nellie, parent=None):
"""
Initializes the NellieAnalysis class.
Parameters
----------
napari_viewer : napari.viewer.Viewer
Reference to the napari viewer instance.
nellie : object
Reference to the main Nellie object containing image and pipeline data.
parent : QWidget, optional
Optional parent widget (default is None).
"""
super().__init__(parent)
self.nellie = nellie
self.viewer = napari_viewer
Expand Down Expand Up @@ -80,6 +171,9 @@ def __init__(self, napari_viewer: 'napari.viewer.Viewer', nellie, parent=None):
self.initialized = False

def reset(self):
"""
Resets the internal state, including the DataFrames and initialization flags.
"""
self.initialized = False
self.voxel_df = None
self.node_df = None
Expand All @@ -88,6 +182,9 @@ def reset(self):
self.image_df = None

def post_init(self):
"""
Initializes UI elements such as checkboxes, buttons, and connects them to their respective event handlers.
"""
self.log_scale_checkbox = QCheckBox("Log scale")
self.log_scale_checkbox.stateChanged.connect(self.on_log_scale)

Expand Down Expand Up @@ -144,6 +241,9 @@ def post_init(self):
self.initialized = True

def set_ui(self):
"""
Sets up the user interface layout, including dropdowns for attribute selection, histogram controls, and export buttons.
"""
main_layout = QVBoxLayout()

# Attribute dropdown group
Expand Down Expand Up @@ -195,6 +295,9 @@ def set_ui(self):
self.setLayout(main_layout)

def _create_dropdown_selection(self):
"""
Creates and configures the dropdown menus for selecting hierarchy levels and attributes.
"""
# Create the dropdown menu
self.dropdown = QComboBox()
self.dropdown.currentIndexChanged.connect(self.on_level_selected)
Expand All @@ -207,17 +310,26 @@ def _create_dropdown_selection(self):
self.set_default_dropdowns()

def set_default_dropdowns(self):
"""
Sets the default values for the hierarchy level and attribute dropdowns.
"""
organelle_idx = self.dropdown.findText('organelle')
self.dropdown.setCurrentIndex(organelle_idx)
area_raw_idx = self.dropdown_attr.findText('organelle_area_raw')
self.dropdown_attr.setCurrentIndex(area_raw_idx)

def check_for_adjacency_map(self):
"""
Checks whether an adjacency map exists, and enables the overlay button if found.
"""
self.overlay_button.setEnabled(False)
if os.path.exists(self.nellie.im_info.pipeline_paths['adjacency_maps']):
self.overlay_button.setEnabled(True)

def rewrite_dropdown(self):
"""
Updates the hierarchy level dropdown based on the available data, and checks for adjacency maps.
"""
self.check_for_adjacency_map()

self.dropdown.clear()
Expand All @@ -234,6 +346,9 @@ def rewrite_dropdown(self):
self.adjacency_maps = None

def export_data(self):
"""
Exports the current graph data as a CSV file to a specified directory.
"""
dt = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
export_dir = self.nellie.im_info.graph_dir
if not os.path.exists(export_dir):
Expand All @@ -254,6 +369,9 @@ def export_data(self):
show_info(f"Data exported to {export_path}")

def save_graph(self):
"""
Saves the current graph as a PNG file to a specified directory.
"""
dt = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
export_dir = self.nellie.im_info.graph_dir
if not os.path.exists(export_dir):
Expand All @@ -267,9 +385,22 @@ def save_graph(self):
show_info(f"Graph saved to {export_path}")

def on_hist_change(self, event):
"""
Event handler for updating the histogram when changes are made (e.g., adjusting the number of bins, min/max values).
"""
self.plot_data(self.dropdown_attr.currentText())

def get_index(self, layer, event):
"""
Retrieves the index of the voxel or feature based on mouse hover coordinates in the napari viewer.
Parameters
----------
layer : Layer
The layer on which the event is triggered.
event : Event
The event triggered by mouse hover.
"""
# get the coordinates of where the mouse is hovering
pos = self.viewer.cursor.position
matched_row = None
Expand Down Expand Up @@ -316,6 +447,9 @@ def get_index(self, layer, event):
self.click_match_table.setVerticalHeaderLabels([f"{t, y, x}\nCSV row"])

def overlay(self):
"""
Applies an overlay of attribute data onto the image in the napari viewer, using adjacency maps for mapping between hierarchy levels.
"""
if self.label_mask is None:
label_mask = self.nellie.im_info.get_memmap(self.nellie.im_info.pipeline_paths['im_instance_label'])
self.label_mask = (label_mask > 0).astype(float)
Expand Down Expand Up @@ -446,24 +580,46 @@ def overlay(self):
self.viewer.reset_view()

def on_t_change(self, event):
"""
Event handler for timepoint changes in the napari viewer. Updates the attribute data and refreshes the plot accordingly.
"""
if self.match_t:
self.on_attr_selected(self.dropdown_attr.currentIndex())

def toggle_match_t(self, state):
"""
Toggles timepoint matching and updates the graph and data accordingly.
Parameters
----------
state : int
The state of the checkbox (checked or unchecked).
"""
if state == 2:
self.match_t = True
else:
self.match_t = False
self.on_attr_selected(self.dropdown_attr.currentIndex())

def toggle_mean_med(self, state):
"""
Toggles between mean and median views for the histogram plot.
Parameters
----------
state : int
The state of the checkbox (checked or unchecked).
"""
if state == 2:
self.is_median = True
else:
self.is_median = False
self.on_attr_selected(self.dropdown_attr.currentIndex())

def get_csvs(self):
"""
Loads the CSV files containing voxel, node, branch, organelle, and image features into DataFrames.
"""
self.voxel_df = pd.read_csv(self.nellie.im_info.pipeline_paths['features_voxels'])
if os.path.exists(self.nellie.im_info.pipeline_paths['features_nodes']):
self.node_df = pd.read_csv(self.nellie.im_info.pipeline_paths['features_nodes'])
Expand All @@ -475,6 +631,14 @@ def get_csvs(self):
# self.voxel_df_idxs = voxel_df[voxel_df.columns[0]]

def on_level_selected(self, index):
"""
Event handler for when a hierarchy level is selected from the dropdown menu.
Parameters
----------
index : int
The index of the selected item in the dropdown.
"""
# This method is called whenever a radio button is selected
# 'button' parameter is the clicked radio button
self.selected_level = self.dropdown.itemText(index)
Expand Down Expand Up @@ -516,6 +680,14 @@ def on_level_selected(self, index):
self.dropdown_attr.addItem(col)

def on_attr_selected(self, index):
"""
Event handler for when an attribute is selected from the dropdown menu.
Parameters
----------
index : int
The index of the selected attribute in the dropdown.
"""
self.hist_reset = True
# if there are no items in dropdown_attr, return
if self.dropdown_attr.count() == 0:
Expand All @@ -541,6 +713,9 @@ def on_attr_selected(self, index):
self.plot_data(selected_attr)

def get_stats(self):
"""
Computes basic statistics (mean, std, median, percentiles) for the currently selected attribute data.
"""
if self.attr_data is None:
return
if not self.log_scale:
Expand All @@ -566,6 +741,9 @@ def get_stats(self):
self.iqr = self.perc75 - self.perc25

def draw_stats(self):
"""
Draws statistics on the histogram plot, including lines for mean, median, std, and percentiles.
"""
if self.attr_data is None:
return
# draw lines for mean, median, std, percentiles on the canvas
Expand All @@ -582,6 +760,14 @@ def draw_stats(self):
self.canvas.draw()

def plot_data(self, title):
"""
Plots the selected attribute data as a histogram, updates the canvas, and displays the computed statistics.
Parameters
----------
title : str
The title for the plot, usually the name of the selected attribute.
"""
self.canvas.figure.clear()
ax = self.canvas.figure.add_subplot(111)
self.data_to_plot = self.data_to_plot.replace([np.inf, -np.inf], np.nan)
Expand Down Expand Up @@ -635,6 +821,14 @@ def plot_data(self, title):
self.hist_reset = False

def on_log_scale(self, state):
"""
Toggles logarithmic scaling for the histogram plot and refreshes the data accordingly.
Parameters
----------
state : int
The state of the checkbox (checked or unchecked).
"""
self.hist_reset = True
if state == 2:
self.log_scale = True
Expand Down

0 comments on commit cc8b206

Please sign in to comment.