added report generation

mtor/dataproc.py

@@ -6,8 +6,20 @@ import pickle
 import seaborn
 import matplotlib.pyplot as pyplot
 
+from reportlab.platypus import (
+    SimpleDocTemplate,
+    Paragraph,
+    PageBreak,
+    Spacer,
+    KeepTogether,
+    Image,
+)
+from reportlab.lib.pagesizes import A4
+from reportlab.lib.styles import getSampleStyleSheet
+from reportlab.lib.units import mm
 from scipy.signal import savgol_filter
 
+
 from .commons import ROI_STATISTIC_FUNCTIONS
 
 
@@ -394,6 +406,119 @@ def save_data(data_frame, selected_df, extremas_df, parameters):
     writer.save()
 
 
+def create_report(data_frame, selected_df, extremas_df, parameters):
+    styles = getSampleStyleSheet()
+    style_headline = styles["Heading1"]
+    style_section = styles["Heading2"]
+    style_text = styles["Normal"]
+
+    data_dir = parameters.data_dir
+    path = data_dir / "report.pdf"
+    doc = SimpleDocTemplate(str(path), pagesize=A4)
+
+    img_width = doc.width * 0.9
+    img_height = (900 * img_width / 1200) * 0.9
+
+    num_images = len(data_frame)
+    num_selected = len(selected_df)
+    num_discarded = num_images - num_selected
+
+    def text_and_graph(text, name):
+        flowable = KeepTogether(
+            [
+                Paragraph(text, style_text),
+                Image(
+                    str(data_dir / name), width=img_width, height=img_height
+                ),
+                Spacer(1, 7 * mm),
+            ]
+        )
+        return flowable
+
+    story = [
+        Paragraph(f"Analysis of {num_images} Tif Images", style_headline),
+        Spacer(1, 10 * mm),
+        Paragraph("Estimating Guard Threshold", style_section),
+        text_and_graph(
+            (
+                "In a first step, the histogram of the combined left and "
+                "right guard values is calculated."
+            ),
+            "1-histogram-of-guard-avarages-not-filtered.png",
+        ),
+        text_and_graph(
+            (
+                "A Savitzky-Golay filter is applied to the histogram to "
+                "smooth the curve."
+            ),
+            "2-histogram-of-guard-avarages-filtered.png",
+        ),
+        text_and_graph(
+            (
+                "The first minima after the first peak is used as the guard "
+                f"threshold value: {int(parameters.guard_max_value)} au"
+            ),
+            "3-histogram-of-guard-avarages-filtered-with-first-minima.png",
+        ),
+        text_and_graph(
+            (
+                "The images with one of the guard values above the threshold "
+                "are discarded."
+            ),
+            "4-image-selection-based-on-guard-values.png",
+        ),
+        Image(
+            str(data_dir / "5-selected-values-based-on-guard-values.png"),
+            width=img_width,
+            height=img_height,
+        ),
+        PageBreak(),
+        Paragraph("Removing Outliers", style_section),
+        text_and_graph(
+            "From the remaining values, outliers are removed.",
+            "6-boxplot-of-guarded-values.png",
+        ),
+        text_and_graph(
+            (
+                f"From {num_images} images {num_discarded} images were "
+                f"discarded, leaving {num_selected} selected. The finally "
+                "selected values are listed in the excel sheet 'selection' "
+                "in the data file."
+            ),
+            "7-selected-images-outliers-removed.png",
+        ),
+        PageBreak(),
+        Paragraph(
+            "Experimental: Applying a rolling min calculation", style_section
+        ),
+        text_and_graph(
+            (
+                "Due to the nature of the experiment, unusable images tend "
+                "to have a higher value as the desiered ones. Therfore a "
+                "rolling min filter is applied"
+            ),
+            "8-selected-images-outliers-removed-rolling-min-applied.png",
+        ),
+        Paragraph("Experimental: Finding Maxima and Minima", style_section),
+        text_and_graph(
+            "To smooth the resulting curve, a Savitzky-Golay filter is used.",
+            (
+                "9-selected-images-outliers-"
+                "removed-rolling-min-savgol-filtered.png"
+            ),
+        ),
+        text_and_graph(
+            (
+                "The most interesting data points should be the maxima and "
+                "minima of this curve. These are listed in the sheet "
+                "'extremas' in the data file"
+            ),
+            "11-finding-minimas.png",
+        ),
+    ]
+    doc.build(story)
+
+
 def save_temp(data_frame, parameters):
     csv_path = parameters.tif_dir / "_data.csv"
     data_frame.to_csv(csv_path, sep="\t")

mtor/workflows.py

@@ -18,6 +18,7 @@ from .dataproc import (
     smooth_savgol_filter,
     find_extremas,
     save_data,
+    create_report,
 )
 from .postproc import (
     stem_file_list,
@@ -28,7 +29,7 @@ from .postproc import (
 
 
 def prescan_workflow(folder, top, right, bottom, left, **kargs):
-    print("1/4: scanning tifs for common autocontrast values")
+    print("1/5: scanning tifs for common autocontrast values")
 
     parameters = Parameters(folder, top, right, bottom, left)
     parameters = scan_tifs(parameters)
@@ -43,7 +44,7 @@ def prescan_workflow(folder, top, right, bottom, left, **kargs):
 
 
 def image_workflow(parameters):
-    print("2/4: Image analysis and conversion")
+    print("2/5: Image analysis and conversion")
     func = functools.partial(process_one_tif, parameters=parameters)
     tif_list = parameters.tif_list
     num_items = len(tif_list)
@@ -56,7 +57,7 @@ def image_workflow(parameters):
 def data_workflow(stats_results, parameters):
     from .dataproc import save_temp
 
-    print("3/4: Data analysis")
+    print("3/5: Data analysis")
     set_plotting_styles()
     data_frame = construct_data_frame(stats_results, parameters)
     save_temp(data_frame, parameters)
@@ -72,22 +73,24 @@ def data_workflow(stats_results, parameters):
     extremas_df = find_extremas(selected_df, parameters)
 
     save_data(data_frame, selected_df, extremas_df, parameters)
+    create_report(data_frame, selected_df, extremas_df, parameters)
 
     return WorkflowResult(list(selected_df["file"]), parameters)
 
 
 def postprocessing_workflow(selected_files, parameters):
-    print("4/4: Post processing")
+    print("4/5: Post processing, sorting cut images")
     file_stems = stem_file_list(selected_files)
-    annotate_color_coded_images(file_stems, parameters)
     sort_cut_images(file_stems, parameters)
     remove_cuts_dir(parameters)
+    print("5/5: Post processing, annotating color coded images")
+    annotate_color_coded_images(file_stems, parameters)
 
 
 def cached_data_workflow(folder):
     from .dataproc import load_temp
 
-    print("3/4: Data analysis")
+    print("3/5: Data analysis")
     set_plotting_styles()
     data_frame, parameters = load_temp(folder)
     find_guard_threshold(data_frame, parameters)
@@ -101,5 +104,6 @@ def cached_data_workflow(folder):
     extremas_df = find_extremas(selected_df, parameters)
 
     save_data(data_frame, selected_df, extremas_df, parameters)
+    create_report(data_frame, selected_df, extremas_df, parameters)
 
     return WorkflowResult(list(selected_df["file"]), parameters)

test_mtor.py

@@ -5,5 +5,5 @@ tif_dir = "original_tifs"
 #mtor.process_tifs(tif_dir, 50, 910, 300, 660, boost=5)
 
 
-mtor.run()
-#mtor.run_data()
+#mtor.run()
+mtor.run_data()