[Demo] Add lollipop chart to ViViVo #874
Conversation
View the example dashboards of the current commit live on PyCafe ☕ 🚀Updated on: 2024-11-15 10:42:05 UTC Link: vizro-core/examples/dev/ Link: vizro-core/examples/scratch_dev |
|
|
||
|
|
||
| @capture("graph") | ||
| def lollipop(data_frame: pd.DataFrame, x: str, y: str): |
There was a problem hiding this comment.
Old code for reference
There was a problem hiding this comment.
The new code is much better 🥇
…/vizro into demo/add-lollipop-chart
| # Unlike the column_and_line chart, where all traces hold equal significance, here the traces differ in importance. | ||
| # The primary scatter plot is the main trace, while the additional traces merely serve as connecting lines. | ||
| # Therefore, should we apply the kwargs solely to the main scatter plot, as illustrated below? | ||
| fig = px.scatter(data_frame, **kwargs) |
There was a problem hiding this comment.
One question: In this case, there is one main scatter trace and several subsidiary scatter traces that only add lines. I assume in this case, we only add kwargs to the main trace and we don't enable the the kwargs for the subsidiary scatter traces?
There was a problem hiding this comment.
I assume in this case, we only add kwargs to the main trace and we don't enable the the kwargs for the subsidiary scatter traces?
I think this is totally correct 👍.
for more information, see https://pre-commit.ci
| ) | ||
|
|
||
| fig.update_layout( | ||
| showlegend=False, yaxis_showgrid=yaxis_showgrid, xaxis_showgrid=xaxis_showgrid, yaxis_rangemode="tozero" |
There was a problem hiding this comment.
I'm wondering if we should generally include layout customizations that optimize a given chart but might not be applicable depending on the data. For instance, setting yaxis_rangemode="tozero" works well for positive data as it removes the gap between the x-axis zeroline and the start of the y-axis, but it's not suitable for negative data.
Additionally, I've just realized that our chart code might not be dynamic enough to handle negative data. Specifically, the calculation code from lines 343-350 does not account for negative values. This might also be the case with other charts, which led me to the question: Does it have to? What's our minimum requirement?
I think we need to refine the minimum requirements a code snippet for the ViViVo needs to satisfy, which is different from a chart that lives in vizro.charts (tbd whether we actually do that). Some of these include:
Potential requirements
- Needs to handle both wide and long form data (Yes)
- Needs to handle both orientations (This depends on the chart type for me - for the lollipop I would have said yes, for the diverging stacked bar chart I would have said: nice to have, but not really necessary)
- Needs to handle negative data
- Needs to be optimized for data viz best practices (this degree can also vary)
- ...
Personally, I see the visual-vocabulary as a guide for best practices in data visualization. I don't mind if some aspects are hard-coded and if the chart doesn't handle all scenarios, as it's meant to be a starting code snippet that users can extend. Trying to handle all of these scenarios will add complexity and reminds me of old VizX times, where we had a bunch of if/else inside the chart code dealing with all possible scenarios that might not be relevant to all users.
huong-li-nguyen
requested review from
stichbury,
Joseph-Perkins,
antonymilne,
maxschulz-COL and
lingyielia
as code owners
| # Unlike the column_and_line chart, where all traces hold equal significance, here the traces differ in importance. | ||
| # The primary scatter plot is the main trace, while the additional traces merely serve as connecting lines. | ||
| # Therefore, should we apply the kwargs solely to the main scatter plot, as illustrated below? | ||
| fig = px.scatter(data_frame, **kwargs) |
There was a problem hiding this comment.
I assume in this case, we only add kwargs to the main trace and we don't enable the the kwargs for the subsidiary scatter traces?
I think this is totally correct 👍.
|
|
||
|
|
||
| @capture("graph") | ||
| def lollipop(data_frame: pd.DataFrame, x: str, y: str): |
There was a problem hiding this comment.
The new code is much better 🥇
Co-authored-by: Petar Pejovic <[email protected]>
…/vizro into demo/add-lollipop-chart
Description
Based on #787 - thank you again yonkmanjl! ⭐
px.scatterScreenshot
Notice
I acknowledge and agree that, by checking this box and clicking "Submit Pull Request":