-
Notifications
You must be signed in to change notification settings - Fork 2.7k
F# samples #7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
F# samples #7
Conversation
System.Environment.SetEnvironmentVariable("Path", System.Environment.GetEnvironmentVariable "Path" + ";" + nativeDirectory) | ||
|
||
module Datasets = | ||
let private buildPath file = System.IO.Path.Combine(@"..\datasets\", file) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm just wondering why this path is pointing to just one folder up instead of two folders up like ..\..\datasets\
. Currently I'm struggling a bit because compiler is looking for dataset
in getting-started
directory. But maybe I'm doing something wrong.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I also had an issue with the path for the data, but in my case it was looking for dataset
in the directory above the repository, so the fix was to use .\datasets\
rather than ..\datasets\
. I assume this is nothing to do with the sample code but rather which directory is current when FSI is launched?
Otherwise, works great.
@robosek @markpattison Yes, I think this is the issue - where you open your IDE. For me, I used VS Code and opened the IDE in the fsharp-samples folder. Evidently, if you open in a different location, the paths don't quite work! Perhaps I'll just change it to assume you're in the root of the repo? |
Oh right @isaacabraham the solution is to open just |
@robosek great idea. |
@robosek @markpattison let me know if it's now any better. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great work! Thanks for make it easy to play ML.NET with FSI 👍
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good.
@OliaG it would be great if you could give everyone interested in this an update on what's needed to get this PR accepted - thanks! |
@isaacabraham thank you very much for your contribution! We'd like to make changes in ML.NET that would provide better support for F#. We are planing to do it in the next coming version v 0.3. So this PR would probably need some updates. After that we'll gladly add it to the repo. |
@OliaG with 0.3 released, is the plumbing now in place for me to update this PR? |
@isaacabraham we did a small reorg, now it will be @CESARDELATORRE or @asthana86 to answer this question. |
@isaacabraham @OliaG - Thanks a lot Isaac for your efforts and PR. You can test v0.3 with your code and see what's new in it in the following blog post we published here: About merging the PR into the repo, it is pretty possible that we'll make significant breaking changes in the API in upcoming versions, so it might be better if we hold on this PR (and additional samples) just for a while. In any case, please tell us when you have tested the code with v0.3, ok? Again, thanks a lot for your PR, it'll be a great help for the community. 👍 |
@CESARDELATORRE Reading through the release notes, I'm not sure that anything has changed since v0.2 i.e. is there support for records (or even class properties)? And is there better support for in-memory data? If neither, then there's not much that can be updated since those are the two main blockers at the moment. Cheers |
@isaacabraham - As far as I know, there are no specific features in v0.3 with improvements for F#. |
@CESARDELATORRE , @isaacabraham Yep we are starting to work with @dsyme to get the story going for F# and ML.NET. Happy to include you on the conversations :). |
@CESARDELATORRE I think @asthana86 is aware of them :-) Probably the biggest two blockers are dotnet/machinelearning#180 and dotnet/machinelearning#91 |
Closing in favour of #36 |
This is a simple port of the Sentiment Analysis sample to F#. This also puts in place the necessary plumbing for other people to easily add new sample scripts (and full-blown apps if required).
Happy to help you get up and running if needed, although there's a readme in the folder with some basic instructions.