Proposal for plugin analytics

setup.py

@@ -49,5 +49,6 @@
         'tensorflow-macos;  platform_system=="Darwin" and platform_machine=="arm64"',
         "napari>=0.4.13",
         "magicgui>=0.3.0",
+        "plausible-events",
     ],
 )

stardist_napari/_analytics.py

@@ -0,0 +1,112 @@
+import platform
+
+import csbdeep
+import magicgui
+import napari
+import numpy as np
+import stardist
+import tensorflow
+from csbdeep.utils.tf import keras_import
+from stardist.models import StarDist2D, StarDist3D
+
+from . import __version__
+
+keras = keras_import()
+
+
+try:
+    from plausible_events import PlausibleEvents
+
+    PE = PlausibleEvents(domain="stardist-napari")
+except:
+    PE = None
+
+
+def consent_event(consent):
+    if PE is None:
+        return
+    PE.event("Consent", dict(value=consent))
+
+
+def launch_event():
+    if PE is None:
+        return
+    # TODO: check for gpus?
+    # gpus = len(tensorflow.config.list_physical_devices('GPU'))
+    props = {
+        "platform": platform.platform().strip(),
+        "python": platform.python_version(),
+        "stardist-napari": __version__,
+    }
+    props.update(
+        {
+            p.__name__: p.__version__
+            for p in (
+                napari,
+                magicgui,
+                tensorflow,
+                keras,
+                csbdeep,
+                stardist,
+            )
+        }
+    )
+    PE.event("Launch", props)
+
+
+def run_event(
+    model,
+    model_selected,
+    models_reg,
+    models_reg_public,
+    x_shape,
+    axes,
+    perc_low,
+    perc_high,
+    norm_image,
+    input_scale,
+    n_tiles,
+    prob_thresh,
+    nms_thresh,
+    cnn_output,
+    norm_axes,
+    output_type,
+    timelapse,
+):
+    if PE is None:
+        return
+
+    def _model_name():
+        # only disclose model names of "public" registered/pre-trained models
+        model_type, model_name = model_selected
+        if model_type in models_reg:
+            return (
+                model_name
+                if (model_name in models_reg_public.get(model_type, {}))
+                else "Custom (registered)"
+            )
+        else:
+            return "Custom (folder)"
+
+    def _shape_pow2(shape, axes):
+        return tuple(
+            s if a == "C" else int(2 ** np.ceil(np.log2(s))) for s, a in zip(shape, axes)
+        )
+
+    props = {
+        "model": _model_name(),
+        "image_shape": _shape_pow2(x_shape, axes),
+        "image_axes": axes,
+        "image_norm": (perc_low, perc_high) if norm_image else False,
+        "image_scale": input_scale,
+        "image_tiles": n_tiles,
+        "thresh_prob": prob_thresh,
+        "thresh_nms": nms_thresh,
+        "output_type": output_type,
+        "output_cnn": cnn_output,
+        "norm_axes": norm_axes,
+    }
+    if "T" in axes:
+        props["timelapse"] = timelapse
+    name = {StarDist2D: "Run 2D", StarDist3D: "Run 3D"}[type(model)]
+    PE.event(name, props)

stardist_napari/_dock_widget.py

@@ -54,7 +54,11 @@ def surface_from_polys(polys):
 def plugin_wrapper():
     # delay imports until plugin is requested by user
     # -> especially those importing tensorflow (csbdeep.models.*, stardist.models.*)
-    from csbdeep.models.pretrained import get_model_folder, get_registered_models
+    from csbdeep.models.pretrained import (
+        get_model_details,
+        get_model_folder,
+        get_registered_models,
+    )
     from csbdeep.utils import (
         _raise,
         axes_check_and_normalize,
@@ -102,7 +106,7 @@ def wrapper(*args):
     # -------------------------------------------------------------------------
 
     _models, _aliases = {}, {}
-    models_reg = {}
+    models_reg, models_reg_public = {}, {}
     for cls in (StarDist2D, StarDist3D):
         # get available models for class
         _models[cls], _aliases[cls] = get_registered_models(cls)
@@ -111,6 +115,15 @@ def wrapper(*args):
             ((_aliases[cls][m][0] if len(_aliases[cls][m]) > 0 else m), m)
             for m in _models[cls]
         ]
+        # keys of registered (i.e. pre-trained) models that are "public",
+        # i.e. transmitting their names for analytics is safe (not personally identifiable)
+        models_reg_public[cls] = [
+            key
+            for name, key in models_reg[cls]
+            if get_model_details(cls, key)[2]["url"].startswith(
+                "https://github.com/stardist/stardist-models/releases/"
+            )
+        ]
 
     model_configs = dict()
     model_threshs = dict()
@@ -136,6 +149,10 @@ def get_model(model_type, model):
 
     # -------------------------------------------------------------------------
 
+    from ._analytics import consent_event, launch_event, run_event
+
+    # -------------------------------------------------------------------------
+
     class Output(Enum):
         Labels = "Label Image"
         Polys = "Polygons / Polyhedra"
@@ -281,6 +298,15 @@ class TimelapseLabels(Enum):
             text="Set optimized postprocessing thresholds (for selected model)",
         ),
         defaults_button=dict(widget_type="PushButton", text="Restore Defaults"),
+        label_analytics=dict(
+            widget_type="Label",
+            label="<b>Analytics:</b>",
+        ),
+        analytics=dict(
+            widget_type="CheckBox",
+            text="Share anonymous usage data",
+            value=False,
+        ),
         progress_bar=dict(label=" ", min=0, max=0, visible=False),
         layout="vertical",
         persist=True,
@@ -312,6 +338,8 @@ def plugin(
         cnn_output,
         set_thresholds,
         defaults_button,
+        label_analytics,
+        analytics,
         progress_bar: mw.ProgressBar,
     ) -> List[napari.types.LayerDataTuple]:
 
@@ -347,7 +375,7 @@ def plugin(
             )  # relevant axes present in input image
             assert len(axes_norm) > 0
             # always jointly normalize channels for RGB images
-            if ("C" in axes and image.rgb == True) and ("C" not in axes_norm):
+            if ("C" in axes and image.rgb is True) and ("C" not in axes_norm):
                 axes_norm = axes_norm + "C"
                 warn("jointly normalizing channels of RGB input image")
             ax = axes_dict(axes)
@@ -370,6 +398,31 @@ def plugin(
             #         _axis = tuple(i for i in range(x.ndim) if i not in (ax['C'],))
             x = normalize(x, perc_low, perc_high, axis=_axis)
 
+        if analytics:
+            try:
+                run_event(
+                    model,
+                    model_selected,
+                    models_reg,
+                    models_reg_public,
+                    x.shape,
+                    axes,
+                    perc_low,
+                    perc_high,
+                    norm_image,
+                    input_scale,
+                    n_tiles,
+                    prob_thresh,
+                    nms_thresh,
+                    cnn_output,
+                    norm_axes,
+                    output_type={t.value: t.name for t in Output}[output_type],
+                    timelapse={t.value: t.name for t in TimelapseLabels}[timelapse_opts],
+                )
+            except Exception as e:
+                if DEBUG:
+                    raise e
+
         # TODO: progress bar (labels) often don't show up. events not processed?
         if "T" in axes:
             app = use_app()
@@ -616,6 +669,7 @@ def progress(it, **kwargs):
     plugin.n_tiles.value = DEFAULTS["n_tiles"]
     plugin.input_scale.value = DEFAULTS["input_scale"]
     plugin.label_head.value = '<small>Star-convex object detection for 2D and 3D images.<br>If you are using this in your research please <a href="https://github.com/stardist/stardist#how-to-cite" style="color:gray;">cite us</a>.</small><br><br><tt><a href="https://stardist.net" style="color:gray;">https://stardist.net</a></tt>'
+    plugin.label_analytics.value = '<small>Help improve this plugin by sharing usage data.<br></small><a href="https://github.com/stardist/stardist-napari/pull/12" style="color:gray;">What is shared and why.</a>'
 
     # make labels prettier (https://doc.qt.io/qt-5/qsizepolicy.html#Policy-enum)
     for w in (plugin.label_head, plugin.label_nn, plugin.label_nms, plugin.label_adv):
@@ -1179,6 +1233,16 @@ def restore_defaults():
         for k, v in DEFAULTS.items():
             getattr(plugin, k).value = v
 
+    # analytics
+    @change_handler(plugin.analytics, init=True)
+    def _analytics_change():
+        widgets_inactive(plugin.label_analytics, active=(not plugin.analytics.value))
+        try:
+            consent_event(plugin.analytics.value)
+        except Exception as e:
+            if DEBUG:
+                raise e
+
     # -------------------------------------------------------------------------
 
     # allow some widgets to shrink because their size depends on user input
@@ -1188,6 +1252,7 @@ def restore_defaults():
     plugin.timelapse_opts.native.setMinimumWidth(240)
 
     plugin.label_head.native.setOpenExternalLinks(True)
+    plugin.label_analytics.native.setOpenExternalLinks(True)
     # make reset button smaller
     # plugin.defaults_button.native.setMaximumWidth(150)
 
@@ -1207,4 +1272,11 @@ def restore_defaults():
     # for i in range(layout.count()):
     #     print(i, layout.itemAt(i).widget())
 
+    if plugin.analytics.value:
+        try:
+            launch_event()
+        except Exception as e:
+            if DEBUG:
+                raise e
+
     return plugin