Dtale.show() error with pyarrow backed pandas dataframe?

[Valence-Shinei]

Greetings, I'm pretty new to Github, but I've been using Dtale for probably around a year now. I recently updated Dtale and Pandas to newer version to take advantage of the pyarrow datatypes in Pandas. So far so good, but when I try to use Dtale.show() as I've done previously, I encounter:

TypeError: 'ArrowExtensionArray' with dtype int64[pyarrow] does not support reduction 'skew'

Attached is the quick setup I did to test this with some quick and dirty dummy data. Please let me know if I should provide more information to help determine if this is an issue. Thank you in advance for your time and consideration of my comments!

[Valence-Shinei]

However, no problems are observed using the standard pandas dataframe with numpy backed datatypes

[Valence-Shinei]

What's also interesting is that Apache Arrow has an article about extension arrays in R, but not for Python 🤔

https://arrow.apache.org/docs/r/reference/new_extension_type.html

[aschonfeld]

Thanks for the detailed ticket @Valence-Shinei so I think what I can do (assuming I can't get the skew function to work) is that I'll just turn off the skew calculation for dtypes that don't have it.

[Valence-Shinei]

You're certainly welcome and if possible and not asking too much, that would be great!

Dtale has really helped me move from Excel into Python in an easier, familiar fashion as I'm learning more Python to upskill, so thank you and team for this amazing library 😊🤝💙.

On Thu, Nov 2, 2023, 11:56 Andrew Schonfeld @.***> wrote:

Thanks for the detailed ticket @Valence-Shinei https://github.com/Valence-Shinei so I think what I can do (assuming I can't get the skew function to work) is that I'll just turn off the skew calculation for dtypes that don't have it.

— Reply to this email directly, view it on GitHub https://github.com/man-group/dtale/issues/823#issuecomment-1791118154, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDIZ45BRZTHJGNFYZ6MA2Q3YCPGFBAVCNFSM6AAAAAA62IUSQOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJRGEYTQMJVGQ . You are receiving this because you were mentioned.Message ID: @.***>

[aschonfeld]

@Valence-Shinei just released v3.8.0 to pypi (will be on conda-forge shortly) with this update. Let me know if it works. Also, if you haven't already would you mind putting your ⭐ on this repo? Thanks 🙏

[Valence-Shinei]

Certainly, I will review and I'll be glad to add my star to the repo 🤝.

My apologies, I'm new to github and this system, but I do want to help give credit and praise where it is due 😎🙌!

On Mon, Nov 27, 2023, 14:09 Andrew Schonfeld @.***> wrote:

@Valence-Shinei https://github.com/Valence-Shinei just released v3.8.0 to pypi (will be on conda-forge shortly) with this update. Let me know if it works. Also, if you haven't already would you mind putting your ⭐ on this repo? Thanks 🙏

— Reply to this email directly, view it on GitHub https://github.com/man-group/dtale/issues/823#issuecomment-1828511342, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDIZ45A3XASDMPKCT6VEBJTYGTXRBAVCNFSM6AAAAAA62IUSQOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMRYGUYTCMZUGI . You are receiving this because you were mentioned.Message ID: @.***>

[Valence-Shinei]

Greetings Andrew, Unfortunately it looks like I'm receiving the same error =/ . Please see the attached screenshots indicating the error after the upgrade.

[aschonfeld]

@Valence-Shinei i'm sorry, for some reason I thought the function didn't exist. I didn't realize the exception is buried within the skew function. I'll release a hotfix handling this. Will keep you posted

[aschonfeld]

@Valence-Shinei just released v3.8.1 let me know how it goes

[Valence-Shinei]

Success =] ! Bravo and thank you Andrew and team 😁🤝!