Data explorer series

https://www.notion.so/qim-dtu/Add-load-directory-to-Data-Explorer-92bbaac99cb54f4d921536f5f3f57e4f?pvs=4

Solution

• Added a checkmark for loading a series
  • If checkmark is selected a textbox will appear. The textbox can then be filled out with the common part of the filename.
• CHANGES: made the loader prioritise the data explorer if a file is selected else it will use the path in the textbox.

qim3d/gui/data_explorer.py

@@ -17,6 +17,7 @@ app.launch()
 
 import datetime
 import os
+import re
 
 import gradio as gr
 import matplotlib
@@ -119,13 +120,13 @@ class Interface:
                                 value="⟳", elem_classes="btn-html h-36"
                             )
                     explorer = gr.FileExplorer(
-                        ignore_glob="*/.*", # ignores hidden files
+                        ignore_glob="*/.*",  # ignores hidden files
                         root_dir=os.getcwd(),
                         label=os.getcwd(),
                         render=True,
                         file_count="single",
                         interactive=True,
-                        elem_classes="h-256 hide-overflow",
+                        elem_classes="h-320 hide-overflow",
                     )
 
                 with gr.Column(scale=1):
@@ -137,22 +138,33 @@ class Interface:
                         interactive=True,
                     )
 
-
-                    virtual_stack = gr.Checkbox(value=False, label="Virtual stack", info="If checked, will use less memory by loading the images on demand.")
-                    load_series = gr.Checkbox(value=False, label="Load series", info="If checked, will load the whole series of images in the same folder as the selected file.")
+                    virtual_stack = gr.Checkbox(
+                        value=False,
+                        label="Virtual stack",
+                        info="If checked, will use less memory by loading the images on demand.",
+                    )
+                    load_series = gr.Checkbox(
+                        value=False,
+                        label="Load series",
+                        info="If checked, will load the whole series of images in the same folder as the selected file.",
+                    )
                     series_contains = gr.Textbox(
                         label="Specify common part of file names for series",
                         value="",
                         visible=False,
                     )
-                    
+
                     dataset_name = gr.Textbox(
                         label="Dataset name (in case of H5 files, for example)",
                         value="exchange/data",
                     )
 
-                    def toggle_show(checkbox): 
-                        return gr.update(visible=True) if checkbox else gr.update(visible=False)
+                    def toggle_show(checkbox):
+                        return (
+                            gr.update(visible=True)
+                            if checkbox
+                            else gr.update(visible=False)
+                        )
 
                     # Show series_contains only if load_series is checked
                     load_series.change(toggle_show, load_series, series_contains)
@@ -211,7 +223,7 @@ class Interface:
                 # Intensity histogram
                 with gr.Column(visible=False) as result_intensity_histogram:
                     hist_plot = gr.Plot(label="Volume intensity histogram")
-                
+
                 # Text box with data summary
                 with gr.Column(visible=False) as result_data_summary:
                     data_summary = gr.Text(
@@ -259,7 +271,6 @@ class Interface:
                 min_projection_plot,
                 hist_plot,
                 data_summary,
-
             ]
 
             ### Listeners
@@ -300,7 +311,7 @@ class Interface:
         # Starts a new session dictionary
         session = Session()
         # Tells rest of the pipeline if the session failed at some point prior, so should skip the rest
-        session.failed = False 
+        session.failed = False
         session.all_operations = Interface().operations
         session.operations = args[0]
         session.base_path = args[1]
@@ -315,20 +326,35 @@ class Interface:
         session.series_contains = args[10]
 
         if session.load_series and session.series_contains == "":
-            session.failed = True
-            raise gr.Error("For series, common part of file name must be provided in 'series_contains' field.")
+
+            # Try to guess the common part of the file names
+            try:
+                filename = session.explorer.split("/")[-1]  # Extract filename from path
+                series_contains = re.search(r"[^0-9]+", filename).group()
+                gr.Info(f"Using '{series_contains}' as common file name part for loading.")
+                session.series_contains = series_contains
+
+            except:
+                session.failed = True
+                raise gr.Error(
+                    "For series, common part of file name must be provided in 'series_contains' field."
+                )
 
         # Get the file path from the explorer or base path
         # priority is given to the explorer if file is selected
         # else the base path is used
-        if session.explorer and (os.path.isfile(session.explorer) or session.load_series): 
+        if session.explorer and (
+            os.path.isfile(session.explorer) or session.load_series
+        ):
             session.file_path = session.explorer
-        elif session.base_path and (os.path.isfile(session.base_path) or session.load_series):
+        elif session.base_path and (
+            os.path.isfile(session.base_path) or session.load_series
+        ):
             session.file_path = session.base_path
         else:
             session.failed = True
             raise gr.Error("Invalid file path")
-        
+
         # If we are loading a series, we need to get the directory
         if session.load_series:
             session.file_path = os.path.dirname(session.file_path)
@@ -436,7 +462,7 @@ class Pipeline:
         # skip loading if session failed at some point prior
         if session.failed:
             return session
-        
+
         try:
             session.vol = load(
                 session.file_path,
@@ -444,8 +470,16 @@ class Pipeline:
                 dataset_name=session.dataset_name,
                 contains=session.series_contains,
             )
+
+            # Incase the data is 4D (RGB for example), we take the mean of the last dimension
+            if session.vol.ndim == 4:
+                session.vol = np.mean(session.vol, axis=-1)
+
+            # The rest of the pipeline expects 3D data
             if session.vol.ndim != 3:
-                raise gr.Error(f"Invalid data shape should be 3 dimensional, not shape: {session.vol.shape}")
+                raise gr.Error(
+                    f"Invalid data shape should be 3 dimensional, not shape: {session.vol.shape}"
+                )
         except Exception as error_message:
             raise gr.Error(
                 f"Failed to load the image: {error_message}"
@@ -473,7 +507,7 @@ class Pipeline:
         # skip loading if session failed at some point prior
         if session.failed:
             return []
-        
+
         self.session = session
         outputs = []
         log.info(session.all_operations)
@@ -652,12 +686,15 @@ class Pipeline:
 
     def plot_vol_histogram(self):
 
-        # The Histogram needs results from the projections    
+        # The Histogram needs results from the projections
         if not self.session.projections_calculated:
             _ = self.get_projections(self.session.vol)
-        
+
         vol_hist, bin_edges = self.vol_histogram(
-            self.session.vol, self.session.nbins, self.session.min_value, self.session.max_value
+            self.session.vol,
+            self.session.nbins,
+            self.session.min_value,
+            self.session.max_value,
         )
 
         fig, ax = plt.subplots(figsize=(6, 4))
@@ -690,11 +727,12 @@ class Pipeline:
 
         return vol_hist, bin_edges
 
-def run_interface(host = "0.0.0.0"):
+
+def run_interface(host="0.0.0.0"):
     gradio_interface = Interface().create_interface()
-    internal_tools.run_gradio_app(gradio_interface,host)
+    internal_tools.run_gradio_app(gradio_interface, host)
 
 
 if __name__ == "__main__":
     # Creates interface
-    run_interface()
\ No newline at end of file
+    run_interface()