joshmarshall1

This PR closes #1611 Adds the basic `register` and `attach` capabilities to DataFrame: `register(name)` `attach(name)` `unregister()` `is_registered()` `unregister_dataframe_by_name(name)` Also adds `.objtype` property to `SegArray` to support using the objtype to...

With the addition of Register and Attach functionality to DataFrame we should explore if it's possible to integrate DataFrame into the Generic Attach workflow

While updating GroupBy to have register and attach functionality, I found that in the case of `Categorical`, `unique_keys.categories` shares the same object as `keys.categories` so during registration the `unique_keys.categories` overwrites...

When registering a SegArray on the server, calling `.unregister()` on the original will remove all components from the symbol table. However, attaching to a registered SegArray then calling `.unregister()` on...

Part of #1793 This issue covers the creation of a .md file to give an overview of how to use the symbol table and our messaging between client and server....

To support the movement of GroupBy to the server it'd be beneficial to first move Categorical to be server-side. This issue tracks moving the functionality and computational aspects of `Categorical`

To support the movement of `GroupBy` to the server it'd be beneficial to first move `Categorical` to be server-side. This issue tracks moving the `__init__` functionality

Add `unregister_by_name` functionality to `SegArray`

Exact problem size will be dependent on machine size, but running both methods with an problem size larger than 10**4 produce an out of memory error Reproducer: ```python N =...