[OTHER] Using chart_config

[JosuaCarl]

What would you like to share or ask?

How can I use chart_config to change the plots in a data_node ? I would like to be able to do scatter plots with only markers, no lines. Is there a way to do this ?
Example:

with tgb.Page(style=style) as visualization:
    tgb.text("## 📊 Data", mode="markdown")
    tgb.data_node("{csv_data_node}", show_owner=True, chart_config={"type": "scatter"})

I don't get how to structure the dictionary, as referenced here: https://docs.taipy.io/en/latest/refmans/reference/pkg_taipy/pkg_gui/pkg_builder/data_node/

Code of Conduct

• I have checked the existing issues.
• I am willing to work on this issue (optional)

[JosuaCarl]

A complete example:

import pandas as pd
import taipy as tp
from taipy.gui import Gui
from taipy.gui import builder as tgb


df = pd.DataFrame(
    {"A": [1,2,3], "B": [4,5,6], "C": [7,8,9]}
)
csv_data_node_config = tp.Config.configure_csv_data_node(
    id="csv_data_node", scope=tp.Scope.GLOBAL
)

csv_data_node = tp.create_global_data_node(csv_data_node_config)
csv_data_node.write(df)

with tgb.Page() as visualization:
    tgb.text("## 📊 Data", mode="markdown")
    tgb.data_node("{csv_data_node}", show_owner=True, chart_config={"type": "scatter"})

gui = Gui(page=visualization)

if __name__ == "__main__":
    gui.run(port=4999)

Clarification of what I want: When I click on the chart view, I want to visualize the data as a different plot (i.e. scatter - markers). Ideally while still being able to choose X and Y, but not necessarily.

[FabienLelaquais]

@JosuaCarl Indeed the chart_config property lacks documentation and examples.
We will handle this quickly. Thanks

[FredLL-Avaiga]

the right property is chart_configs
it is a dict based on datanode config id

[JosuaCarl]

Does this mean there is no general option? So I would have to enter something like:

{"my_data_node":
    {"type": "scatter"}
}

Because this also does nothing.

[FabienLelaquais]

Does this mean there is no general option? So I would have to enter something like:

{"my_data_node":
{"type": "scatter"}
}
Because this also does nothing.

@JosuaCarl: the property name is wrong in the doc (should be chart_configs as @FredLL-Avaiga said).
This property can hold a dictionary where each key is a data node configuration identifier.
The corresponding value indicates Taipy how to deal with a dataset, should it come from a DataNode issued from that DataNodeConfig.
Many of the entries are arrays, because we need to handle multiple traces.
We are in the process of documenting this entirely. We're not quite there yet.

In your case you may want to try setting the types property of the dict set for your data node config id to an array with the desired trace type. Because you only have one trace, that will be easy: ["scatter"].
I have not tried this though.
Please keep us posted, someone will jump on this.

Thx

[FabienLelaquais]

@JosuaCarl
Small addition for you to go forward waiting for the whole thing.

In your case, the use case you came up with is you want chart that displays the the data node data as markers only.
to achieve this, you would create the data node control configuration dictionary as follows:

data_node_viewer_configs = {
    "csv_data_node": {
        "modes": ["markers"]
    }
}

then use it when building your page:

tgb.data_node("{csv_data_node}", show_owner=True, chart_configs=data_node_viewer_configs )

This should accomplish what you wanted.
Please let us know, cheers.