PD.DataFrame is not supported for State

[qu3vipon]

Describe the bug pd.DataFrame type is not supported for State

To Reproduce

class State(pc.State):
   data: pd.DataFrame

Error Message

$ TypeError: State vars must be primitive Python types, Plotly figures, Pandas dataframes, or subclasses of pc.Base. Found var “data” with type <class ‘pandas.core.frame.DataFrame’>.

Code

• couldn’t find pd.DataFrame in Valid state var types

# pynecone/utils.py L60

# Valid state var types
PrimitiveType = Union[int, float, bool, str, list, dict, tuple]
StateVar = Union[PrimitiveType, Base, None]

Specifics (please complete the following information):

• Python Version: 3.8.10
• Pynecone Version: 0.1.12
• OS: mac OS

[Alek99]

Looking into this now

[Alek99]

This is a bug on our side will be fixing asap

[Alek99]

In the mean time it seems computed vars for pandas is working see this example, https://github.com/pynecone-io/pynecone-examples/blob/main/nba/nba/nba.py

[wongni]

In the mean time it seems computed vars for pandas is working see this example, https://github.com/pynecone-io/pynecone-examples/blob/main/nba/nba/nba.py

@Alek99 When I used the computed var in the data_table's data prop I am getting this error. Is this also a bug?

            pc.data_table(
                data=State.df,
                pagination=True,
                search=True,
                sort=True,
                resizable=True,
            ),

[Alek99]

Hmm just ran the example and it seems to be working for me what pynecone version are you on and did you change the code at all

[wongni]

Hmm just ran the example and it seems to be working for me what pynecone version are you on and did you change the code at all

I am using pynecone 0.1.13 version and python 3.9.13 version. And the only code I changed in the nba example's main branch was from data=nba_data, to data=State.df,. I tried both Firefox and Chrome browsers on Mac but still got An error happened while fetching the data error on the table.

[picklelo]

@wongni I'm seeing this too - working on a fix now.

[rohitsathish]

What's the right way to instantiate a pandas dataframe in rx.State? Still seem to be getting errors with this approach.