Samples second pass for Clustering Trainer #3317

artidoro · 2019-04-12T21:22:22Z

Tracked in #2522.

This PR removes the dependency on SampleUtils for clustering trainers samples (KMeans samples).

zeahmed · 2019-04-12T21:39:00Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/Clustering.ttinclude

+                {
+                    Label = (uint)label,
+                    // Create random features with two clusters.
+                    // The first half has feature values cetered around 0.6 the second half has values centered around 0.4.


cetered [](start = 57, length = 7)

centered #Resolved

zeahmed · 2019-04-12T21:42:33Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/KMeans.cs

-            //      The coordinates of centroid 0 are: (26, 6, 1)
+            // Expected output similar to:
+            //   The first 3 coordinates of the first centroid are: (0.6035213, 0.6017533, 0.5964218)
+            //   The first 3 coordinates of the second centroid are: (0.4031044, 0.4175443, 0.4082336)
            //
            // Note: use the advanced options constructor to set the number of threads to 1 for a deterministic behavior.


// Note: use the advanced options constructor to set the number of threads to 1 for a deterministic behavior. [](start = 12, length = 109)

This comment seems out of context. Specifically what does advance options constructor mean here? #Resolved

zeahmed

rogancarr · 2019-04-12T22:19:44Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/Clustering.ttinclude

+            foreach (var p in predictions.Take(2))
+                Console.WriteLine($"Label: {p.Label}, Prediction: {p.PredictedLabel}");
+            foreach (var p in predictions.TakeLast(3))
+                Console.WriteLine($"Label: {p.Label}, Prediction: {p.PredictedLabel}");


Most of the time, clustering doesn't have a label. Maybe we should make one sample for labeled clustering, one for label-free clustering. #Resolved

I added some comments to explain that the Label column is not used during training. This is simply for comparison with the predicted label.

In reply to: 275080292 [](ancestors = 275080292)

rogancarr · 2019-04-12T22:20:13Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/Clustering.ttinclude

+            <#=ExpectedCentroidsOutput#>
+        }
+
+        private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed = 0)


int count [](start = 71, length = 9)

Set this to the default used above. #Resolved

rogancarr · 2019-04-12T22:20:39Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/Clustering.ttinclude

+            <#=ExpectedCentroidsOutput#>
+        }
+
+        private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed = 0)


int seed = 0 [](start = 82, length = 12)

I'm superstitious about passing a 0 to someone else's random number generator. Maybe 1? #Pending

We use 0 in all our templates for the method GenerateRandomDataPoints. I don't see what it should be an issue

In reply to: 275080461 [](ancestors = 275080461)

rogancarr · 2019-04-12T22:21:40Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/Clustering.ttinclude

+        private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed = 0)
+        {
+            var random = new Random(seed);
+            float randomFloat() => (float)random.NextDouble();


(float)random.NextDouble() [](start = 35, length = 26)

- 0.5f #Pending

I don't think that centering at 0 will give any significant advantage.

In reply to: 275080652 [](ancestors = 275080652)

rogancarr · 2019-04-12T22:22:45Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/KMeansWithOptions.tt

+string TrainerOptions = @"KMeansTrainer.Options
+            {
+                NumberOfClusters = 2,
+                MaximumNumberOfIterations = 100,


MaximumNumberOfIterations = 100 [](start = 16, length = 31)

I don't think we should suggest tuning this. Mostly we want this option to not be hit. (I consider things we use in our samples to be suggestions.) #Resolved

rogancarr · 2019-04-12T22:23:26Z

docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Clustering/KMeans.tt

+string ClassName = "KMeans";
+            string Trainer = "KMeans";
+            string TrainerOptions = null;
+            string InlineTrainerOptions = "numberOfClusters: 2";


numberOfClusters: 2 [](start = 43, length = 19)

10

Are you suggesting to use more clusters?

In reply to: 275080967 [](ancestors = 275080967)

shmoradims

codecov · 2019-04-15T23:32:42Z

Codecov Report

❗ No coverage uploaded for pull request base (master@8644b3b). Click here to learn what that means.
The diff coverage is n/a.

@@            Coverage Diff            @@
##             master    #3317   +/-   ##
=========================================
  Coverage          ?   72.69%           
=========================================
  Files             ?      807           
  Lines             ?   145172           
  Branches          ?    16225           
=========================================
  Hits              ?   105539           
  Misses            ?    35220           
  Partials          ?     4413

Flag	Coverage Δ
#Debug	`72.69% <ø> (?)`
#production	`68.23% <ø> (?)`
#test	`88.97% <ø> (?)`

artidoro added the documentation Related to documentation of ML.NET label Apr 12, 2019

artidoro requested review from wschin, shmoradims, rogancarr and zeahmed April 12, 2019 21:22

artidoro self-assigned this Apr 12, 2019

zeahmed reviewed Apr 12, 2019

View reviewed changes

zeahmed approved these changes Apr 12, 2019

View reviewed changes

rogancarr reviewed Apr 12, 2019

View reviewed changes

shmoradims mentioned this pull request Apr 15, 2019

Docs and samples for the API reference site (P0 & P1 Trainers) #2522

Closed

shmoradims approved these changes Apr 15, 2019

View reviewed changes

artidoro force-pushed the clusteringsamples branch from fc2c2b2 to 421e91e Compare April 16, 2019 18:05

artidoro added 2 commits April 16, 2019 16:16

Kmeans no sample utils

27c3350

review comments

0e31b8b

artidoro force-pushed the clusteringsamples branch from 421e91e to 0e31b8b Compare April 16, 2019 23:19

artidoro merged commit 32bd0e2 into dotnet:master Apr 17, 2019

ghost locked as resolved and limited conversation to collaborators Mar 22, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Samples second pass for Clustering Trainer #3317

Samples second pass for Clustering Trainer #3317

Uh oh!

artidoro commented Apr 12, 2019

Uh oh!

zeahmed Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

zeahmed Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

zeahmed left a comment

Uh oh!

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

artidoro Apr 15, 2019

Uh oh!

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

artidoro Apr 15, 2019

Uh oh!

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

artidoro Apr 15, 2019

Uh oh!

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

Uh oh!

rogancarr Apr 12, 2019

Uh oh!

artidoro Apr 15, 2019

Uh oh!

shmoradims left a comment

Uh oh!

codecov bot commented Apr 15, 2019 •

edited

Loading

Uh oh!

Uh oh!

Samples second pass for Clustering Trainer #3317

Samples second pass for Clustering Trainer #3317

Uh oh!

Conversation

artidoro commented Apr 12, 2019

Uh oh!

zeahmed Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zeahmed Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zeahmed left a comment

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

artidoro Apr 15, 2019

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

artidoro Apr 15, 2019

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

artidoro Apr 15, 2019

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019 • edited by artidoro Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

rogancarr Apr 12, 2019

Choose a reason for hiding this comment

Uh oh!

artidoro Apr 15, 2019

Choose a reason for hiding this comment

Uh oh!

shmoradims left a comment

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Apr 15, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

zeahmed Apr 12, 2019 •

edited by artidoro

Loading

zeahmed Apr 12, 2019 •

edited by artidoro

Loading

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

rogancarr Apr 12, 2019 •

edited by artidoro

Loading

codecov bot commented Apr 15, 2019 •

edited

Loading