AnswerDotAI / AnswerDotAI/nbdev

Dataframe HTML tags showing up in exported README

Offen
#1,155 5 Kommentare 0 Reaktionen 0 zugewiesene Personen Auf GitHub ansehen
bug
Vorherrschende Sprache
Jupyter Notebook
Sterne
5.3k
Forks
513
Ø Merge
2 T. 30 Min.
Gemergte PRs (30 T.)
8

Beschreibung

Related to https://github.com/fastai/nbdev/issues/474

Rendered readme on Github: https://github.com/rxavier/poniard/blob/experiment/nbdev/README.md
index.ipynb: https://github.com/rxavier/poniard/blob/experiment/nbdev/nbs/index.ipynb

Basically, "raw output" dataframes (for example, when a method simply returns a dataframe, like `get_results` in this example) get this added before the actual table in markdown:

```
.dataframe tbody tr th:only-of-type { vertical-align: middle; }
.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

```

Meanwhile, dataframes rendered with `display(HTML(df.to_html()))` (which shows after `setup` in this example) display fine.

Quarto HTML docs render any of these fine.

Edit: I implemented an ugly temporary fix by using directives and ipython's `display`:

```python
#| output: false

obj.method_that_outputs_a_df()
```

```python
#| echo: false

from ipython import display, HTML

df = obj.method_that_outputs_a_df()
display(HTML(df.to_html()))
```

Beitragsleitfaden

Beitragsleitfaden öffnen

Rechercherichtung

Start with nbs/index.ipynb and the exported README linked in the issue, then reproduce a raw dataframe return such as get_results and compare it with display(HTML(df.to_html())). The fix is complete when exported README markdown renders the dataframe without the unwanted style tags, while the table remains visible.

Vom Indexierungsmodell aus dem Issue-Text verfasst.

Bewertung

Tech-Stack
jupyter-notebook, python
Bereich
documentation
Issue-Typ
Bug
Schwierigkeit
4/5
Geschätzter Aufwand
3-5 Tage
Aktivitätsstatus
Veraltet
Klarheit
Größtenteils klar
Anfängerfreundlichkeit
35/100

Neue Issues direkt in Ihr Postfach

Eine kurze Übersicht über anfängerfreundliche GitHub-Issues.