Skip to content

Display single text outputs

If you are

  • Developping a new component
  • Testing various inputs on an existing component
  • ...

you might want to quickly apply a pipeline and display the output doc in a comprehensible way.

from edsnlp.viz import QuickExample

E = QuickExample(nlp)  #

Next, simply call E with any string:

txt = "Le patient présente une anomalie."
E(txt)

                              Le patient présente une anomalie                               
                                                                                             
┏━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━┓
┃ Entity   ┃ Source   ┃ eds.hypoth… ┃ eds.negation ┃ eds.family ┃ eds.history ┃ eds.report… ┃
┡━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━┩
│ patient  │ patient  │ False       │ False        │ False      │ False       │ False       │
│ anomalie │ anomalie │ False       │ False        │ False      │ False       │ False       │
└──────────┴──────────┴─────────────┴──────────────┴────────────┴─────────────┴─────────────┘

By default, each Qualifiers in nlp adds a corresponding column to the output. Additionnal informations can be displayed by using the extensions parameter. For instance, if entities have a custom ent._.custom_ext extensions, it can be displayed by providing the extension when instantiating QuickExample:

E = QuickExample(nlp, extensions=["_.custom_ext"])

Finally, if you prefer to output a DataFrame instead of displaying a table, set the as_dataframe parameter to True:

E = QuickExample(nlp)
E(txt, as_dataframe=True)