Documentation samples for binary classifiers (Static API) #1456
Conversation
| /// Downloads the adult train dataset from the ML.NET repo | ||
| /// </summary> | ||
| public static string DownloadAdultTrainDataset() | ||
| => Download("https://github.com/raw/dotnet/machinelearning/master/test/data/adult.train", "adult.train"); |
There was a problem hiding this comment.
"https://github.com/raw/dotnet/machinelearning/master/test/data/adult.trai [](start = 24, length = 85)
"https://github.com/raw/dotnet/machinelearning/master/test/data/adult.trai [](start = 24, length = 85)
It needs to include the hash, otherwise it might change overtime.
| /// Downloads the adult test dataset from the ML.NET repo | ||
| /// </summary> | ||
| public static string DownloadAdultTestDataset() | ||
| => Download("https://github.com/raw/dotnet/machinelearning/master/test/data/adult.test", "adult.test"); |
There was a problem hiding this comment.
https://github.com/raw/dotnet/machinelearning/master/test/data/adult.test" [](start = 25, length = 85)
Similar here:
https://github.com/dotnet/machinelearning/blob/20c36ca769d6b1f4fc816de4ae5f46266e8edd9b/test/data/adult.train
#Resolved
| // It takes care of logging, exception tracking and as a source of randomness | ||
| // Using random seed and automatic level of concurrency | ||
| using (var environment = new ConsoleEnvironment(seed: 0, conc: 0)) | ||
| { |
There was a problem hiding this comment.
this is not necessary, the MLContext below is sufficient.
| row.Ethnicity.OneHotEncoding(), | ||
| row.Sex.OneHotEncoding(), | ||
| row.HoursPerWeek, | ||
| row.NativeCountry.OneHotEncoding().SelectFeaturesBasedOnCount(count: 10)), |
There was a problem hiding this comment.
.SelectFeaturesBasedOnCount(count: 10)) [](start = 66, length = 39)
@shmoradims to check whether this is necessary.
|
Thanks for the contribution, @bojanmisic. This looks great, a few small comments. If you sync to latest, you'll notice that we are now doing one sample per file, to avoid those pesky line changes in the codebase :) |
bojanmisic
force-pushed
the
docs_for_binary_classifiers_static
branch
2 times, most recently
from
5360763 to
7bb2625
Compare
bojanmisic
force-pushed
the
docs_for_binary_classifiers_static
branch
from
4a85dff to
2ecbcf1
Compare
|
|
||
| // NOTE: WHEN ADDING TO THE FILE, ALWAYS APPEND TO THE END OF IT. | ||
| // If you change the existing content, check that the files referencing it in the XML documentation are still correct, as they reference | ||
| // line by line. |
There was a problem hiding this comment.
❔ Does it not have the ability to import a named region? #Resolved
There was a problem hiding this comment.
I am actually going to remove the copyright and change the references to import the whole file.
In reply to: 231153004 [](ancestors = 231153004)
bojanmisic
changed the title
|
@sfilipi Thank you for the review, I have updated the branch per your suggestions. Also updated the PR description to reflect the things I have changed. Removed WIP. |
ghost
locked as resolved and limited conversation to collaborators
Addresses #1257.
Adding documentation samples for binary classifiers when using
Static API:SDCAFastTreeLightGBMAveragedPerceptronA couple of points I'd like to mention:
adult.traindataset both for training and testing (90/10 split) in the examples (sinceadult.testis not properly formatted).FastTreein both Regression and Classification contexts, I had to rename already added examples fromFastTree.cstoFastTreeRegression.csand made sure the change is reflected in the docs so the examples point to the correct files.