Skip to content
This repository has been archived by the owner on Aug 27, 2020. It is now read-only.

add compat to dependencies #10

Closed
kafisatz opened this issue Jun 4, 2019 · 5 comments
Closed

add compat to dependencies #10

kafisatz opened this issue Jun 4, 2019 · 5 comments

Comments

@kafisatz
Copy link

kafisatz commented Jun 4, 2019

When using your package, I get an error saying that you should add Compat to your dependencies:

julia> using FastGroupBy
[ Info: Precompiling FastGroupBy [e6aeac8e-85c6-520e-8612-effc59bc0f33]
┌ Warning: Package FastGroupBy does not have Compat in its dependencies:
│ - If you have FastGroupBy checked out for development and have
│   added Compat as a dependency but haven't updated your primary
│   environment's manifest file, try `Pkg.resolve()`.
│ - Otherwise you may need to report an issue with FastGroupBy
└ Loading Compat into FastGroupBy from project dependency, future warnings for FastGroupBy are suppressed.
@xiaodaigh
Copy link
Owner

can you try ]dev FastGroupBy. I don't' get the same error in the latest dev version on Julia 1.1.1.

@kafisatz
Copy link
Author

kafisatz commented Jun 9, 2019

I am on 1.0.4 for various reasons. And will stay on that version for a while (I think)

@xiaodaigh
Copy link
Owner

I am on 1.0.4 for various reasons

Is ]dev FastGroupBy an option for you? I might look at get a new release soon

@kafisatz
Copy link
Author

kafisatz commented Jun 9, 2019

Thanks that does actually work for me.

However, I noted that there is some additional functionality which your package does currently not offered (compared to DataFrames.aggregate and DataFrames.groupby) which are:

  1. Considering your last example in the Readme: if I have several :grps, say I want to aggregate by [:grps1, :grps2, :grps3] that does not seem possible in this package.
  2. if :grps is a custom type which I have defined (for which I have defined isless and thus sort). FastBy does not seem to work. I realize that I may miss some (or all) of the speedup of this package in this case.

Thank you for the support though.

@xiaodaigh
Copy link
Owner

Done with Compas deps.

As to mulitple groups that was planned but I couldn't get it to run faster than R's data.table. I've tried so many approaches. I will slowly add them. Thanks.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants