1 Labels

1.1 Creating a Region Label

Labels can be used to create annotations in the GUI. You can create your own label, which must inherit BaseRegionLabel, like this:

from mad_gui.plot_tools.base_label import BaseRegionLabel
from mad_gui import start_gui

class Status(BaseRegionLabel):
   # This label will always be shown at the lowest 20% of the plot view
   min_height = 0
   max_height = 0.2
   name = "Anomaly Label"

   # Snapping will be done, if you additionally pass a Settings object to the GUI,
   # which has an attribute SNAP_AXIS. See the README, the part of Adjusting Constants
   # for more information
   snap_to_min = False
   # snap_to_max = False  # if setting this to `True`, set `snap_to_min` to `False` or delete it

   # User will be asked to set the label's description when creating a label.
   # This can have an arbitrary amount of levels with nested dictionaries.
   # This es an example for a two-level description:
   descriptions = {"normal": None, "anomaly": ["too fast", "too slow"]}

Using your algorithm in the GUI

You need to pass the label class (and optionally other plugins) to the start_gui function, see Readme: Developing Plugins.

The description defines the possible strings that can be assigned to a label. They will automatically show up after adding a new label or by clicking on a label when in Edit label mode, such that the user can select one of the descriptions. In our exemplary video, this is {"stand": None, "walk": ["fast", "slow"], "jump": None}.

1.2 Creating an Event Label

This kind of events can be plotted in the GUI, when the user uses Ctrl + LeftMouseClick/Space. See the above section for more infos on class attributes, and our exemplary video for more information on labelling.

class MyEvent(BaseEventLabel):
    min_height = 0
    max_height = 1
    name = "Peak"
    descriptions = {"Positive peak": None, "Negative peak": None}
    snap_to_min = False
    snap_to_max = False

1.3 Using more than one category of labels

You can also pass several labels to the GUI. See the image below the code snippet to find out, what kind of annotations can be created with these label classes:

 from mad_gui.plot_tools.labels.base_label import BaseRegionLabel
 from mad_gui import start_gui

 class LayerOne(BaseRegionLabel):
    # This label will always be shown at the upper 20% of the plot view
    min_height = 0.8
    max_height = 1
    name = "Activity"
    # description = see first code snippet on this page

 class LayerTwo(BaseRegionLabel):
    # This label will always be shown at the lowest 20% to 80% of the plot view
    min_height = 0.2
    max_height = 0.8
    name = "Sub-Activity"
    # description = see first code snippet on this pag

start_gui(labels=[LayerOne, LayerTwo])