Dashboard¶
Class and code for rendering the ipywidgets dashboard.
-
class
tx2.dashboard.
Dashboard
(transformer_wrapper: tx2.wrapper.Wrapper, show_umap=True, show_salience=True, show_word_count=True, show_cluster_salience=True, show_cluster_sample_btns=True, show_wordclouds=False, show_scoring=True, point_size=5, unfocused_point_size=20, highlighted_point_size=75)¶ Class that handles the setup, visualizations, and customization of the ipywidgets dashboard.
Methods:
__init__
(transformer_wrapper[, show_umap, …])Constructor.
render
()Return combined layout widget
Attributes:
The hex RGB colors to use in the scatter plot and word frequency graphs to represent each possible category.
Size to render highlighted and currently selected points in the UMAP plot.
Size to render points in the UMAP plot.
Show the top aggregate-salience “important” words per cluster graphs.
Show the sampling buttons for each cluster.
Show the current entry salience map “heatmap”.
Show aggregate scoring metrics, confusion matrices, etc.
Show the UMAP projection/embedding scatter plot.
Show the per-cluster top word frequency graphs.
Show the wordclouds for each cluster.
The
tx2.wrapper.Wrapper
instance/interface into the transformer.Size to render unfocused background points in the UMAP plot.
-
__init__
(transformer_wrapper: tx2.wrapper.Wrapper, show_umap=True, show_salience=True, show_word_count=True, show_cluster_salience=True, show_cluster_sample_btns=True, show_wordclouds=False, show_scoring=True, point_size=5, unfocused_point_size=20, highlighted_point_size=75)¶ Constructor.
- Parameters
transformer_wrapper – the
tx2.wrapper.Wrapper
instance/interface into the transformer.show_umap – Show the UMAP projection/embedding scatter plot.
show_salience – Show the current entry salience map “heatmap”.
show_word_count – Show the per-cluster top word frequency graphs.
show_cluster_salience – Show the top aggregate-salience “important” words per cluster graphs.
show_cluster_sample_btns – Show the sampling buttons for each cluster.
show_wordclouds – Show the wordclouds for each cluster.
show_scoring – Show aggregate scoring metrics, confusion matrices, etc.
point_size – Size to render points in the UMAP plot.
unfocused_point_size – Size to render unfocused background points in the UMAP plot.
selected_point_size – Size to render highlighted and currently selected points in the UMAP plot.
-
colors
¶ The hex RGB colors to use in the scatter plot and word frequency graphs to represent each possible category. By default this is set to the ‘Category20’ palette. The number of clusters cannot exceed the number of colors in this array. (Currently it’s automatically limited to the first 20 clusters.)
-
highlighted_point_size
¶ Size to render highlighted and currently selected points in the UMAP plot.
-
point_size
¶ Size to render points in the UMAP plot.
-
render
()¶ Return combined layout widget
-
show_cluster_salience
¶ Show the top aggregate-salience “important” words per cluster graphs.
-
show_cluster_sample_btns
¶ Show the sampling buttons for each cluster.
-
show_salience
¶ Show the current entry salience map “heatmap”.
-
show_scoring
¶ Show aggregate scoring metrics, confusion matrices, etc.
-
show_umap
¶ Show the UMAP projection/embedding scatter plot.
-
show_word_count
¶ Show the per-cluster top word frequency graphs.
-
show_wordclouds
¶ Show the wordclouds for each cluster.
-
transformer_wrapper
¶ The
tx2.wrapper.Wrapper
instance/interface into the transformer.
-
unfocused_point_size
¶ Size to render unfocused background points in the UMAP plot.
-