autogaita 1.4.0__tar.gz → 1.4.1__tar.gz

{autogaita-1.4.0 → autogaita-1.4.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogaita
-Version: 1.4.0
+Version: 1.4.1
 Summary: Automatic Gait Analysis in Python. A toolbox to streamline and standardise the analysis of kinematics across species after ML-based body posture tracking. Despite being optimised for gait analyses, AutoGaitA has the potential to be used for any kind of kinematic analysis.
 Home-page: https://github.com/mahan-hosseini/AutoGaitA/
 Author: Mahan Hosseini

{autogaita-1.4.0 → autogaita-1.4.1}/README.md

@@ -7,11 +7,12 @@
 [![paper: biorxiv](https://img.shields.io/badge/paper-biorxiv-blue)](https://doi.org/10.1101/2024.04.14.589409) 
 
 ![Black](https://img.shields.io/badge/code%20style-black-000000.svg)
+[![Tutorials](https://img.shields.io/badge/YouTube-red?style=for-the-badge&logo=youtube&logoColor=white)](https://youtube.com/playlist?list=PLCn5T7K_H8K56NIcEsfDK664OP7cN_Bad&feature=shared)
 [![X URL](https://img.shields.io/twitter/url?url=https%3A%2F%2Fx.com%2Fautogaita&style=social&label=updates)](https://x.com/autogaita)
 
 # Automated Gait Analysis in Python 🐸
 
-- AutoGaitA simplifies, accelerates, and standardises gait analyses (as well as the analysis of other rhythmic behaviours) after body posture tracking in 2D with [DeepLabCut](https://github.com/DeepLabCut/DeepLabCut) and [SLEAP](https://github.com/talmolab/sleap) or marker-based as well as marker-less (such as [Simi Motion](http://www.simi.com/en/products/movement-analysis/simi-motion-2d3d.html?type=rss%2F)) methods for obtaining 3D coordinates. 
+- AutoGaitA simplifies, accelerates, and standardises gait analyses (as well as the analysis of other rhythmic behaviours) after body posture tracking in 2D with [DeepLabCut](https://github.com/DeepLabCut/DeepLabCut) and [SLEAP](https://github.com/talmolab/sleap) or marker-based as well as marker-less methods for obtaining 3D coordinates. 
 - AutoGaitA's first-level tools provide a wide range of automated kinematic analyses for each input video and AutoGaitA Group allows the comparison of up to six groups. 
 - AutoGaitA enables comparisons to be made across experimental conditions, species, disease states or genotypes. 
 - Despite being developed with gait data, AutoGaitA can be utilised for the analysis of any motor behaviour.
@@ -46,7 +47,7 @@ It is strongly recommended that a separate virtual environment for AutoGaitA is
 
 **[The AutoGaitA YouTube Channel](https://youtube.com/playlist?list=PLCn5T7K_H8K56NIcEsfDK664OP7cN_Bad&si=mV5p2--nYvbofkPh) provides tutorials for file preparation and instructions on how to use AutoGaitA. This includes in-depth explanations of all details, (main & advanced) configurations, possibilities, and outputs.**
 
-*Please note that tutorial videos might not always reflect the most up-to-date version of our toolbox, especially in the beginning when things are regularly changing. We will make sure to record new videos whenever there are major changes though. Last tutorial-update was with v0.4.0. (August 2024).*
+*Please note that tutorial videos might not always reflect the most up-to-date version of our toolbox, especially in the beginning when things are regularly changing. We will make sure to record new videos whenever there are major changes though. If updates are just minor we refer to them in the notes below each video. Last tutorial-update was with v1.4.0. (September 2025).*
 
 ### Example Data
 We provide an example dataset in the **example data** folder of this repository, with a set of mice walking over differently wide beams and both the beam as well as body coordinates being tracked with DLC. Note that this dataset was used in our tutorial videos introducing *AutoGaitA DLC*, *AutoGaitA Group* and in our video explaining file preparation for *AutoGaitA DLC*.  We further provide a **group** folder there that can be used alongside the *AutoGaitA Group* tutorial to confirm that users generate the same set of results following our instructions.
@@ -68,9 +69,9 @@ Users are advised to read the **General Recommendations** section of that folder
 By default, *AutoGaitA DLC* and *AutoGaitA Universal 3D* implement standard values for mouse and human locomotion, respectively. If your analysis deviates from these standards (e.g. by focussing on another behaviour or a different species) **you must change these values!** 
 
 **Find out more about *AutoGaitA's custom joints and angles:***
-- [YouTube - AutoGaitA DLC Advanced Configuration](https://youtu.be/MP9g9kXRE_Q?feature=shared) 
-- [YouTube - AutoGaitA Universal 3D (prev. called Simi)](https://youtu.be/rTG-Fc9XI9g?feature=shared) 
-- [Documentation - AutoGaitA DLC](https://docs.google.com/document/d/1iQxSwqBW3VdIXHm-AtV4TGlgpJPDldogVx6qzscsGxA/edit?tab=t.0#heading=h.20bg7b7ymt0b)
+- [YouTube - AutoGaitA DLC/SLEAP Advanced Configuration](https://youtu.be/MP9g9kXRE_Q?feature=shared) 
+- [YouTube - AutoGaitA Universal 3D](https://youtu.be/vZyWmPcI6ao) 
+- [Documentation - AutoGaitA DLC/SLEAP](https://docs.google.com/document/d/1iQxSwqBW3VdIXHm-AtV4TGlgpJPDldogVx6qzscsGxA/edit?tab=t.0#heading=h.20bg7b7ymt0b)
 - [Documentation - AutoGaitA Universal 3D](https://docs.google.com/document/d/1iQxSwqBW3VdIXHm-AtV4TGlgpJPDldogVx6qzscsGxA/edit?tab=t.0#heading=h.uz61bpmua7qz)
 
 ### Bin number of step cycle normalisation
@@ -95,7 +96,7 @@ AutoGaitA's cfg files and dictionaries sometimes change as a result, which means
 We document each version's cfg-changes in [AutoGaitA Releases](https://github.com/mahan-hosseini/AutoGaitA/releases), which is particularly relevant for users wrapping custom scripts around AutoGaitA's functions.
 
 ## Reference
-If you use this code or data please [cite our preprint](https://www.biorxiv.org/content/10.1101/2024.04.14.589409v1).
+If you use this code or data please [cite our preprint](https://doi.org/10.1101/2024.04.14.589409).
 
 ## License
 AutoGaitA is licensed under [GPL v3.0](https://github.com/mahan-hosseini/AutoGaitA/blob/main/LICENSE) and Forschungszentrum Jülich GmbH holds all copyrights. 
@@ -106,6 +107,8 @@ The AutoGaitA software is provided without warranty of any kind, express or impl
 [Mahan Hosseini](https://github.com/mahan-hosseini)
 
 ## Contributors
+[Sarah Sabbagh](https://github.com/sarahhsabbagh) - Undergraduate Student
+
 [Luca Flemming](https://github.com/luca-flemming) - Undergraduate Student
 
 [Nicholas del Grosso](https://github.com/nickdelgrosso) - RSE Advisor

{autogaita-1.4.0 → autogaita-1.4.1}/autogaita/batchrun_scripts/group_dlcrun.py

@@ -15,22 +15,18 @@ def group_dlcrun():
         "5mm",
         "12mm",
         "25mm",
-        # Interaction Issue Dataset
-        # "Control",
-        # "Silenced",
         # Testing Dataset
         # "one",
         # "two",
+        # 6 Group Dataset
+        # "36w_5mm",
+        # "36w_12mm",
+        # "36w_25mm",
+        # "104w_5mm",
+        # "104w_12mm",
+        # "104w_25mm",
     ]
     folderinfo["group_dirs"] = [
-        # --------------------------
-        # Full Dataset (39 subjects)
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Sanity Test Interaction Graziana/Result 10 joints/Control Full",
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Sanity Test Interaction Graziana/Result 10 joints/Silenced Full/",
-        # --------------------------
-        # Reduced Dataset (5 subjects)
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Sanity Test Interaction Graziana/Result 10 joints/Control",
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Sanity Test Interaction Graziana/Result 10 joints/Silenced/",
         # --------------------------
         # Testing Dataset
         # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Testing/Group 1/",
@@ -40,17 +36,23 @@ def group_dlcrun():
         "/Users/mahan/sciebo/PythonCode/gaita_repo/example data/5mm/Results/",
         "/Users/mahan/sciebo/PythonCode/gaita_repo/example data/12mm/Results/",
         "/Users/mahan/sciebo/PythonCode/gaita_repo/example data/25mm/Results/",
+        # --------------------------
+        # 6 Group Dataset (nothing needed)
     ]
     folderinfo["results_dir"] = (
-        "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/example_data_results/"
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Sanity Test Interaction Graziana/Result 10 joints/Mahan Results/"
-        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Testing/GroupResults/"
+        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/example_data_results/"
+        "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Testing/GroupResults/"
+        # "/Users/mahan/Desktop/Test6GroupResults/"
+    )
+    folderinfo["load_dir"] = (
+        ""
+        # "/Users/mahan/sciebo/Research/AutoGaitA/Mouse/Final Mouse Results/AdultElder3BeamsForPaper"  # for 6 group analysis example
     )
-    folderinfo["load_dir"] = ""
     # cfg
     cfg = {}
-    cfg["do_permtest"] = False  # True
+    cfg["do_permtest"] = True
     cfg["do_anova"] = True
+    cfg["anova_design"] = "RM ANOVA"
     cfg["permutation_number"] = 100
     cfg["PCA_n_components"] = 6
     # cfg["PCA_n_components"] = 0.33
@@ -59,14 +61,11 @@ def group_dlcrun():
     cfg["PCA_save_3D_video"] = False  # True
     cfg["PCA_bins"] = ""  # "0-10,24,50-75"
     cfg["stats_threshold"] = 0.05
-    cfg["plot_SE"] = False
-    cfg["color_palette"] = "viridis"
+    cfg["plot_SE"] = True
+    cfg["color_palette"] = "Set2"
     cfg["dont_show_plots"] = True
     cfg["legend_outside"] = True
     cfg["which_leg"] = "left"
-    cfg["anova_design"] = "RM ANOVA"
-    # cfg["anova_design"] = "Mixed ANOVA"
-    cfg["permutation_number"] = 100
     cfg["PCA_variables"] = [
         # "Hind paw tao y",
         # "Ankle y",
@@ -84,14 +83,14 @@ def group_dlcrun():
         # "Hip Angle",
     ]
     cfg["stats_variables"] = [
-        "Hind paw tao y",
-        # "Ankle y",
-        "Knee y",
+        # "Hind paw tao y",
+        "Ankle y",
+        # "Knee y",
         # "Ankle Angle",
-        "Knee Angle",
+        # "Knee Angle",
         # "Nose x",
         # "Knee y",
-        "Knee x",
+        # "Knee x",
         # "Knee Velocity",
         # "Knee Acceleration",
         # "Knee Angle",

{autogaita-1.4.0 → autogaita-1.4.1}/autogaita/common2D/common2D_utils.py

@@ -16,7 +16,14 @@ def run_singlerun_in_multirun(tracking_software, idx, info, folderinfo, cfg):
     this_info = {}
     keynames = info.keys()
     for keyname in keynames:
-        this_info[keyname] = info[keyname][idx]
+        # if the key is leading mouse/run num zeros, only pass it if it is not False
+        # => we want these to not be keys in this_info if it is False due to how
+        #    leading zeros are handled in the prep script
+        if "leading_" in keyname:
+            if info[keyname][idx] is not False:
+                this_info[keyname] = info[keyname][idx]
+        else:  # pass as is for all other keys
+            this_info[keyname] = info[keyname][idx]
         if cfg["results_dir"]:
             this_info["results_dir"] = os.path.join(
                 cfg["results_dir"], this_info["name"]
@@ -51,7 +58,13 @@ def extract_info(tracking_software, folderinfo, in_GUI=False):
     elif tracking_software == "SLEAP":
         file_type = ".h5"
     # prepare output info dict and run
-    info = {"name": [], "mouse_num": [], "run_num": []}
+    info = {
+        "name": [],
+        "mouse_num": [],
+        "run_num": [],
+        "leading_mouse_num_zeros": [],
+        "leading_run_num_zeros": [],
+    }
     for filename in os.listdir(folderinfo["root_dir"]):
         # make sure we don't get wrong files
         if (
@@ -109,14 +122,10 @@ def extract_info(tracking_software, folderinfo, in_GUI=False):
                     info["name"].append(this_name)
                     info["mouse_num"].append(this_mouse_num)
                     info["run_num"].append(this_run_num)
-    # if we had to fix leading zeros, make sure to save them so we can use them in
-    # move_data_to_folders function later
-    # => make sure it's lists of strings because this is needed by the
-    #    run_singlerun_in_multiruns function later
-    if leading_mouse_num_zeros:
-        info["leading_mouse_num_zeros"] = [leading_mouse_num_zeros]
-    if leading_run_num_zeros:
-        info["leading_run_num_zeros"] = [leading_run_num_zeros]
+                    # note that leading_zeros are always appended but only transfered
+                    # to this_info dicts in singlerun_from_multi if not False
+                    info["leading_mouse_num_zeros"].append(leading_mouse_num_zeros)
+                    info["leading_run_num_zeros"].append(leading_run_num_zeros)
     # this might happen if user entered wrong identifiers or folder
     if len(info["name"]) < 1:
         no_files_message = (
@@ -159,6 +168,8 @@ def find_number(fullstring, prestring, poststring):
     while fullstring[start_idx] == "0" and start_idx < end_idx - 1:
         start_idx += 1
         leading_zeros += "0"
+    if leading_zeros == "":
+        leading_zeros = False  # important that it is False if none present
     return int(fullstring[start_idx:end_idx]), leading_zeros
 
 

{autogaita-1.4.0 → autogaita-1.4.1}/autogaita/group/group_4_stats.py

@@ -5,6 +5,7 @@ from autogaita.group.group_utils import (
     save_figures,
     ytickconvert_mm_to_cm,
     ylabel_velocity_and_acceleration,
+    setup_stats_plots_vars,
 )
 import os
 import sys
@@ -38,8 +39,6 @@ from autogaita.group.group_constants import (
     MULTCOMP_RESULT_P_IDENTIFIER,
     MULTCOMP_RESULT_SPLIT_STR,
     MULTCOMP_EXCEL_COLS,
-    STATS_PLOT_LEGEND_SIZE,  # PLOTS
-    STATS_PLOTS_SUPLABEL_SIZE,
     BOX_COLOR,
     BOX_ALPHA,
     STD_ALPHA,
@@ -324,12 +323,10 @@ def plot_permutation_test_results(
     dont_show_plots = cfg["dont_show_plots"]
     legend_outside = cfg["legend_outside"]
 
-    if len(contrasts) > 3:  # if we have 4 groups or more, N/2x2 subplot layout
-        f, ax = plt.subplots(int(round(len(contrasts) / 2)), 2, layout="constrained")
-        ax = ax.ravel()
-    else:
-        f, ax = plt.subplots(len(contrasts), 1, layout="constrained")
-    x = np.linspace(0, 100, bin_num)
+    # call preparation function for vars and loop over contrasts to plot subplots
+    f, ax, stats_plots_legend_size, stats_plots_suplabel_size, x = (
+        setup_stats_plots_vars(contrasts, bin_num)
+    )
     for c, contrast in enumerate(contrasts):
         # prepare group strings and (importantly!) index of current groups from _NAMES
         groups = [group_name for group_name in contrast.split(CONTRAST_SPLIT_STR)]
@@ -347,82 +344,46 @@ def plot_permutation_test_results(
             else:
                 std = g_std_dfs[group_idx].iloc[:, y_col]
             this_color = group_color_dict[group_name]
-            if type(ax) == np.ndarray:  # so we can do 2-group contrasts
-                ax[c].plot(x, y, color=this_color, label=group_name, zorder=1)
-                ax[c].fill_between(
-                    x,
-                    y - std,
-                    y + std,
-                    color=this_color,
-                    alpha=STD_ALPHA,
-                    lw=STD_LW,
-                    zorder=1,
-                )
-            else:
-                ax.plot(x, y, color=this_color, label=group_name, zorder=1)
-                ax.fill_between(
-                    x,
-                    y - std,
-                    y + std,
-                    color=this_color,
-                    alpha=STD_ALPHA,
-                    lw=STD_LW,
-                    zorder=1,
-                )
+            # plot line and fill between for error
+            ax[c].plot(x, y, color=this_color, label=group_name, zorder=1)
+            ax[c].fill_between(
+                x,
+                y - std,
+                y + std,
+                color=this_color,
+                alpha=STD_ALPHA,
+                lw=STD_LW,
+                zorder=1,
+            )
         # adjust legend & convert to cm (if needed) before plotting clusters
-        if type(ax) == np.ndarray:
-            if legend_outside is True:
-                ax[c].legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax[c].legend(fontsize=STATS_PLOT_LEGEND_SIZE)
-            if check_mouse_conversion(feature, cfg, stats_var=stats_var):
-                ytickconvert_mm_to_cm(ax[c])
-        else:
-            if legend_outside is True:
-                ax.legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE + 4,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax.legend(fontsize=STATS_PLOT_LEGEND_SIZE + 4)
-            if check_mouse_conversion(feature, cfg, stats_var=stats_var):
-                ytickconvert_mm_to_cm(ax)
+        if legend_outside is True:
+            ax[c].legend(
+                fontsize=stats_plots_legend_size,
+                loc="center left",
+                bbox_to_anchor=(1, 0.5),
+            )
+        elif legend_outside is False:
+            ax[c].legend(fontsize=stats_plots_legend_size)
+        if check_mouse_conversion(feature, cfg, stats_var=stats_var):
+            ytickconvert_mm_to_cm(ax[c])
         # plot significant clusters
         # => note that clusters is a list of list with idxs between 0 & bin_num-1
         clusters = extract_all_clusters(trueobs_results_df, contrast)
-        if type(ax) == np.ndarray:
-            ymin = ax[c].get_ylim()[0]
-            ymax = ax[c].get_ylim()[1]
-        else:
-            ymin = ax.get_ylim()[0]
-            ymax = ax.get_ylim()[1]
+        ymin = ax[c].get_ylim()[0]
+        ymax = ax[c].get_ylim()[1]
         for cluster in x[clusters]:  # index x with clusters == cluster has correct val
             x_coords = [cluster[0], cluster[1], cluster[1], cluster[0]]
             y_coords = [ymin, ymin, ymax, ymax]
-            if type(ax) == np.ndarray:
-                ax[c].fill(
-                    x_coords,
-                    y_coords,
-                    color=BOX_COLOR,
-                    alpha=BOX_ALPHA,
-                    lw=STD_LW,
-                    zorder=0,
-                )
-            else:
-                ax.fill(
-                    x_coords,
-                    y_coords,
-                    color=BOX_COLOR,
-                    alpha=BOX_ALPHA,
-                    lw=STD_LW,
-                    zorder=0,
-                )
-    f.supxlabel("Percentage", fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            ax[c].fill(
+                x_coords,
+                y_coords,
+                color=BOX_COLOR,
+                alpha=BOX_ALPHA,
+                lw=STD_LW,
+                zorder=0,
+            )
+    # plotting is done: now do figure-level stuff (suplabels, save, add to GUI)
+    f.supxlabel("Percentage", fontsize=stats_plots_suplabel_size)
     # ylabels depend on whether we converted mm to cm and on the feature
     # code below calls the ylabel function for all possible cases:
     # 1) (DLC only) converted velocity & acceleration
@@ -434,10 +395,10 @@ def plot_permutation_test_results(
         if feature in ["Velocity", "Acceleration"]:
             f.supylabel(
                 ylabel_velocity_and_acceleration(feature, "x in cm", sampling_rate),
-                fontsize=STATS_PLOTS_SUPLABEL_SIZE,
+                fontsize=stats_plots_suplabel_size,
             )
         else:
-            f.supylabel(feature + " (cm)", fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            f.supylabel(feature + " (cm)", fontsize=stats_plots_suplabel_size)
     else:
         if feature in ["Velocity", "Acceleration"]:
             if "Angle" in stats_var:
@@ -449,14 +410,13 @@ def plot_permutation_test_results(
                     unit = "Y"
             f.supylabel(
                 ylabel_velocity_and_acceleration(feature, unit, sampling_rate),
-                fontsize=STATS_PLOTS_SUPLABEL_SIZE,
+                fontsize=stats_plots_suplabel_size,
             )
         else:
-            f.supylabel(feature, fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            f.supylabel(feature, fontsize=stats_plots_suplabel_size)
     figure_file_string = stats_var + " - Cluster-extent Test"
-    f.suptitle(figure_file_string, fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+    f.suptitle(figure_file_string, fontsize=stats_plots_suplabel_size)
     save_figures(f, results_dir, figure_file_string)
-
     # add figure to plot panel figures list
     if dont_show_plots is False:  # -> show plot panel
         plot_panel_instance.figures.append(f)
@@ -757,8 +717,10 @@ def plot_multcomp_results(
     dont_show_plots = cfg["dont_show_plots"]
     legend_outside = cfg["legend_outside"]
 
-    f, ax = plt.subplots(len(contrasts), 1, layout="constrained")
-    x = np.linspace(0, 100, bin_num)
+    # call preparation function for vars and loop over contrasts to plot subplots
+    f, ax, stats_plots_legend_size, stats_plots_suplabel_size, x = (
+        setup_stats_plots_vars(contrasts, bin_num)
+    )
     for c, contrast in enumerate(contrasts):
         # prepare group strings and (importantly!) index of current groups from _NAMES
         groups = [group_name for group_name in contrast.split(CONTRAST_SPLIT_STR)]
@@ -776,103 +738,56 @@ def plot_multcomp_results(
             else:
                 std = g_std_dfs[group_idx].iloc[:, y_col]
             this_color = group_color_dict[group_name]
-            if type(ax) == np.ndarray:  # so we can do a 2-way contrast
-                ax[c].plot(x, y, color=this_color, label=group_name, zorder=1)
-                ax[c].fill_between(
-                    x,
-                    y - std,
-                    y + std,
-                    color=this_color,
-                    alpha=STD_ALPHA,
-                    lw=STD_LW,
-                    zorder=1,
-                )
-            else:
-                ax.plot(x, y, color=this_color, label=group_name, zorder=1)
-                ax.fill_between(
-                    x,
-                    y - std,
-                    y + std,
-                    color=this_color,
-                    alpha=STD_ALPHA,
-                    lw=STD_LW,
-                    zorder=1,
-                )
+            # plot line and fill between for error
+            ax[c].plot(x, y, color=this_color, label=group_name, zorder=1)
+            ax[c].fill_between(
+                x,
+                y - std,
+                y + std,
+                color=this_color,
+                alpha=STD_ALPHA,
+                lw=STD_LW,
+                zorder=1,
+            )
         # adjust legend & convert to cm (if needed) before plotting clusters
-        if type(ax) == np.ndarray:
-            if legend_outside is True:
-                ax[c].legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax[c].legend(fontsize=STATS_PLOT_LEGEND_SIZE)
-            if check_mouse_conversion(feature, cfg, stats_var=stats_var):
-                ytickconvert_mm_to_cm(ax[c])
-        else:
-            if legend_outside is True:
-                ax.legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE + 4,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax.legend(fontsize=STATS_PLOT_LEGEND_SIZE + 4)
-            if check_mouse_conversion(feature, cfg, stats_var=stats_var):
-                ytickconvert_mm_to_cm(ax)
+        if legend_outside is True:
+            ax[c].legend(
+                fontsize=stats_plots_legend_size,
+                loc="center left",
+                bbox_to_anchor=(1, 0.5),
+            )
+        elif legend_outside is False:
+            ax[c].legend(fontsize=stats_plots_legend_size)
+        if check_mouse_conversion(feature, cfg, stats_var=stats_var):
+            ytickconvert_mm_to_cm(ax[c])
         # plot significant clusters
         clusters = extract_multcomp_significance_clusters(
             multcomp_df, contrast, stats_threshold
         )
-        if type(ax) == np.ndarray:
-            ymin = ax[c].get_ylim()[0]
-            ymax = ax[c].get_ylim()[1]
-        else:
-            ymin = ax.get_ylim()[0]
-            ymax = ax.get_ylim()[1]
+        ymin = ax[c].get_ylim()[0]
+        ymax = ax[c].get_ylim()[1]
         for cluster in x[clusters]:  # index x with clusters == cluster has correct val
             x_coords = [cluster[0], cluster[1], cluster[1], cluster[0]]
             y_coords = [ymin, ymin, ymax, ymax]
-            if type(ax) == np.ndarray:
-                ax[c].fill(
-                    x_coords,
-                    y_coords,
-                    color=BOX_COLOR,
-                    alpha=BOX_ALPHA,
-                    lw=STD_LW,
-                    zorder=0,
-                )
-            else:
-                ax.fill(
-                    x_coords,
-                    y_coords,
-                    color=BOX_COLOR,
-                    alpha=BOX_ALPHA,
-                    lw=STD_LW,
-                    zorder=0,
-                )
-        if type(ax) == np.ndarray:
-            # legend adjustments
-            if legend_outside is True:
-                ax[c].legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax[c].legend(fontsize=STATS_PLOT_LEGEND_SIZE)
-        else:
-            # legend adjustments
-            if legend_outside is True:
-                ax.legend(
-                    fontsize=STATS_PLOT_LEGEND_SIZE + 4,
-                    loc="center left",
-                    bbox_to_anchor=(1, 0.5),
-                )
-            elif legend_outside is False:
-                ax.legend(fontsize=STATS_PLOT_LEGEND_SIZE + 4)
-    f.supxlabel("Percentage", fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            ax[c].fill(
+                x_coords,
+                y_coords,
+                color=BOX_COLOR,
+                alpha=BOX_ALPHA,
+                lw=STD_LW,
+                zorder=0,
+            )
+        # legend adjustments
+        if legend_outside is True:
+            ax[c].legend(
+                fontsize=stats_plots_legend_size,
+                loc="center left",
+                bbox_to_anchor=(1, 0.5),
+            )
+        elif legend_outside is False:
+            ax[c].legend(fontsize=stats_plots_legend_size)
+    # plotting is done: now do figure-level stuff (suplabels, save, add to GUI)
+    f.supxlabel("Percentage", fontsize=stats_plots_suplabel_size)
     # ylabels depend on whether we converted mm to cm and on the feature
     # code below calls the ylabel function for all possible cases:
     # 1) (DLC only) converted velocity & acceleration
@@ -884,10 +799,10 @@ def plot_multcomp_results(
         if feature in ["Velocity", "Acceleration"]:
             f.supylabel(
                 ylabel_velocity_and_acceleration(feature, "x in cm", sampling_rate),
-                fontsize=STATS_PLOTS_SUPLABEL_SIZE,
+                fontsize=stats_plots_suplabel_size,
             )
         else:
-            f.supylabel(feature + " (cm)", fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            f.supylabel(feature + " (cm)", fontsize=stats_plots_suplabel_size)
     else:
         if feature in ["Velocity", "Acceleration"]:
             if "Angle" in stats_var:
@@ -899,17 +814,16 @@ def plot_multcomp_results(
                     unit = "Y"
             f.supylabel(
                 ylabel_velocity_and_acceleration(feature, unit, sampling_rate),
-                fontsize=STATS_PLOTS_SUPLABEL_SIZE,
+                fontsize=stats_plots_suplabel_size,
             )
         else:
-            f.supylabel(feature, fontsize=STATS_PLOTS_SUPLABEL_SIZE)
+            f.supylabel(feature, fontsize=stats_plots_suplabel_size)
     figure_file_string = stats_var + " - Tukey's Multiple Comparison Test"
     f.suptitle(
         figure_file_string,
-        fontsize=STATS_PLOTS_SUPLABEL_SIZE,
+        fontsize=stats_plots_suplabel_size,
     )
     save_figures(f, results_dir, figure_file_string)
-
     # add figure to plot panel figures list
     if dont_show_plots is False:  # -> show plot panel
         plot_panel_instance.figures.append(f)

{autogaita-1.4.0 → autogaita-1.4.1}/autogaita/group/group_constants.py

@@ -48,7 +48,7 @@ PCA_BARPLOT_BARCOLOR = "#c74767"  # deep rose
 PCA_BARPLOT_LINECOLOR = "#005f6a"  # petrol
 PCA_CUSTOM_SCATTER_OUTER_SEPARATOR = ";"
 PCA_CUSTOM_SCATTER_INNER_SEPARATOR = ","
-STATS_PLOT_LEGEND_SIZE = 6
+STATS_PLOTS_LEGEND_SIZE = 6
 STATS_PLOTS_SUPLABEL_SIZE = 12
 BOX_COLOR = "#fe420f"  # significance boxes - col = orangered
 BOX_ALPHA = 0.1

{autogaita-1.4.0 → autogaita-1.4.1}/autogaita/group/group_utils.py

@@ -4,7 +4,11 @@ from autogaita.resources.constants import INFO_TEXT_WIDTH
 from autogaita.group.group_constants import (
     GROUP_CONFIG_TXT_FILENAME,
     PCA_CUSTOM_SCATTER_OUTER_SEPARATOR,
+    STATS_PLOTS_LEGEND_SIZE,
+    STATS_PLOTS_SUPLABEL_SIZE,
 )
+import numpy as np
+import matplotlib.pyplot as plt
 
 
 # %% .........................  print start and finish  ................................
@@ -202,3 +206,28 @@ def check_mouse_conversion(feature, cfg, **kwargs):
                 elif "Angle" not in kwargs["stats_var"]:
                     return True
     return False
+
+
+def setup_stats_plots_vars(contrasts, bin_num):
+    """Set up variables for all stats plots"""
+    if len(contrasts) <= 5:
+        f, ax = plt.subplots(len(contrasts), 1, layout="constrained")
+        stats_plots_suplabel_size = STATS_PLOTS_SUPLABEL_SIZE
+        if len(contrasts) == 1:
+            stats_plots_legend_size = STATS_PLOTS_LEGEND_SIZE + 4
+            ax = np.atleast_1d(ax)  # quick hack so we can index ax as a list
+        else:
+            stats_plots_legend_size = STATS_PLOTS_LEGEND_SIZE
+    else:
+        if len(contrasts) == 6:  # 4 groups
+            f, ax = plt.subplots(3, 2, layout="constrained", figsize=(15, 10))
+        elif len(contrasts) == 10:  # 5 groups
+            f, ax = plt.subplots(5, 2, layout="constrained", figsize=(15, 10))
+        elif len(contrasts) == 15:  # 6 groups
+            f, ax = plt.subplots(5, 3, layout="constrained", figsize=(15, 10))
+        ax = ax.ravel()  # so we can index it as 1D array
+        stats_plots_suplabel_size = STATS_PLOTS_SUPLABEL_SIZE + 5
+        stats_plots_legend_size = STATS_PLOTS_LEGEND_SIZE + 1
+    # x is linear space between 0 and 100% with bin_num steps
+    x = np.linspace(0, 100, bin_num)
+    return f, ax, stats_plots_legend_size, stats_plots_suplabel_size, x