Skip to content

Upgrade all regressors to use TT #3319

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.

Sign up for GitHub

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

Merged
merged 27 commits into from
Apr 16, 2019
Merged

Upgrade all regressors to use TT #3319

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
27 commits
Select commit Hold shift + click to select a range
4020dfd
Add Gam simple samples
wschin Apr 12, 2019
746e323
Add advanced Gam samples back
wschin Apr 12, 2019
2a696f0
Update FastForest samples
wschin Apr 12, 2019
c3da068
Remove items added by accident
wschin Apr 12, 2019
4bb7f85
Update FastTree
wschin Apr 12, 2019
d8c48f4
Replace tab with spaces
wschin Apr 12, 2019
8083107
Rename advanced sample files
wschin Apr 12, 2019
21c4876
Update LightGbm
wschin Apr 12, 2019
c6d6b18
Add FastTreeTweedie and fix format
wschin Apr 12, 2019
6864142
Rename two files
wschin Apr 12, 2019
9c246d1
Update OLS and fix build
wschin Apr 12, 2019
3cfe768
Fix csproj
wschin Apr 12, 2019
06f67b7
Upgrade SDCA
wschin Apr 12, 2019
e64a6df
Merge branch 'master' into tt-reg
wschin Apr 13, 2019
6721259
Add missing links to docs
wschin Apr 13, 2019
02eea47
Add spaces back
wschin Apr 15, 2019
53a2beb
Reduce test set's size
wschin Apr 15, 2019
29b556d
Move class comment to example()
wschin Apr 15, 2019
775c3b6
Fix LightGBM nuget reference
wschin Apr 15, 2019
5ef7b7b
Fix some comments
wschin Apr 16, 2019
8364916
Address comments
wschin Apr 16, 2019
6e038f1
Avoid local function in ttinclude
wschin Apr 16, 2019
952ee58
Rename sample files
wschin Apr 16, 2019
8e6386c
Update file names
wschin Apr 16, 2019
efe91ed
Update OGD doc strings
wschin Apr 16, 2019
04c99cf
Merge branch 'master' into tt-reg
wschin Apr 16, 2019
aadb4ff
Fix csproj
wschin Apr 16, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
47 changes: 28 additions & 19 deletions docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Regression/FastForest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,40 +6,40 @@

namespace Samples.Dynamic.Trainers.Regression
{
public static class FastForest
public static class FastForestRegression
{
// This example requires installation of additional NuGet package
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
public static void Example()
{
// Create a new context for ML.NET operations. It can be used for exception tracking and logging,
// as a catalog of available operations and as the source of randomness.
// Setting the seed to a fixed number in this example to make outputs deterministic.
var mlContext = new MLContext(seed: 0);

// Create a list of training examples.
var examples = GenerateRandomDataPoints(1000);
// Create a list of training data points.
var dataPoints = GenerateRandomDataPoints(1000);

// Convert the examples list to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(examples);
// Convert the list of data points to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(dataPoints);

// Define the trainer.
var pipeline = mlContext.Regression.Trainers.FastForest();
var pipeline = mlContext.Regression.Trainers.FastForest(labelColumnName: nameof(DataPoint.Label), featureColumnName: nameof(DataPoint.Features));

// Train the model.
var model = pipeline.Fit(trainingData);

// Create testing examples. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(500, seed:123));
// Create testing data. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(5, seed: 123));

// Run the model on test data set.
var transformedTestData = model.Transform(testData);

// Convert IDataView object to a list.
var predictions = mlContext.Data.CreateEnumerable<Prediction>(transformedTestData, reuseRowObject: false).ToList();

// Look at 5 predictions
foreach (var p in predictions.Take(5))
// Look at 5 predictions for the Label, side by side with the actual Label for comparison.
foreach (var p in predictions)
Console.WriteLine($"Label: {p.Label:F3}, Prediction: {p.Score:F3}");

// Expected output:
Expand All @@ -51,27 +51,26 @@ public static void Example()

// Evaluate the overall metrics
var metrics = mlContext.Regression.Evaluate(transformedTestData);
Microsoft.ML.SamplesUtils.ConsoleUtils.PrintMetrics(metrics);
PrintMetrics(metrics);

// Expected output:
// Mean Absolute Error: 0.06
// Mean Squared Error: 0.01
// Mean Squared Error: 0.00
// Root Mean Squared Error: 0.07
// RSquared: 0.93
// RSquared: 0.96 (closer to 1 is better. The worest case is 0)
}

private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed=0)
{
var random = new Random(seed);
float randomFloat() => (float)random.NextDouble();
for (int i = 0; i < count; i++)
{
var label = randomFloat();
float label = (float)random.NextDouble();
yield return new DataPoint
{
Label = label,
// Create random features that are correlated with label.
Features = Enumerable.Repeat(label, 50).Select(x => x + randomFloat()).ToArray()
// Create random features that are correlated with the label.
Features = Enumerable.Repeat(label, 50).Select(x => x + (float)random.NextDouble()).ToArray()
};
}
}
Expand All @@ -92,5 +91,15 @@ private class Prediction
// Predicted score from the trainer.
public float Score { get; set; }
}

// Print some evaluation metrics to regression problems.
private static void PrintMetrics(RegressionMetrics metrics)
{
Console.WriteLine($"Mean Absolute Error: {metrics.MeanAbsoluteError:F2}");
Console.WriteLine($"Mean Squared Error: {metrics.MeanSquaredError:F2}");
Console.WriteLine($"Root Mean Squared Error: {metrics.RootMeanSquaredError:F2}");
Console.WriteLine($"RSquared: {metrics.RSquared:F2}");
}
}
}
}

24 changes: 24 additions & 0 deletions docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Regression/FastForest.tt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,24 @@
<#@ include file="RegressionSamplesTemplate.ttinclude"#>

<#+
string ClassHeader = @"// This example requires installation of additional NuGet package
// <a href=""https://www.nuget.org/packages/Microsoft.ML.FastTree/"">Microsoft.ML.FastTree</a>. ";

string ClassName="FastForestRegression";
string ExtraUsing = null;
string Trainer = @"FastForest(labelColumnName: nameof(DataPoint.Label), featureColumnName: nameof(DataPoint.Features))";
string TrainerOptions = null;

string ExpectedOutputPerInstance= @"// Expected output:
// Label: 0.985, Prediction: 0.864
// Label: 0.155, Prediction: 0.164
// Label: 0.515, Prediction: 0.470
// Label: 0.566, Prediction: 0.501
// Label: 0.096, Prediction: 0.138";

string ExpectedOutput = @"// Expected output:
// Mean Absolute Error: 0.06
// Mean Squared Error: 0.00
// Root Mean Squared Error: 0.07
// RSquared: 0.96 (closer to 1 is better. The worest case is 0)";
#>
47 changes: 29 additions & 18 deletions docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Regression/FastForestWithOptions.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,26 +7,28 @@

namespace Samples.Dynamic.Trainers.Regression
{
public static class FastForestWithOptions
public static class FastForestWithOptionsRegression
{
// This example requires installation of additional NuGet package
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
public static void Example()
{
// Create a new context for ML.NET operations. It can be used for exception tracking and logging,
// as a catalog of available operations and as the source of randomness.
// Setting the seed to a fixed number in this example to make outputs deterministic.
var mlContext = new MLContext(seed: 0);

// Create a list of training examples.
var examples = GenerateRandomDataPoints(1000);
// Create a list of training data points.
var dataPoints = GenerateRandomDataPoints(1000);

// Convert the examples list to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(examples);
// Convert the list of data points to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(dataPoints);

// Define trainer options.
var options = new FastForestRegressionTrainer.Options
{
LabelColumnName = nameof(DataPoint.Label),
FeatureColumnName = nameof(DataPoint.Features),
// Only use 80% of features to reduce over-fitting.
FeatureFraction = 0.8,
// Create a simpler model by penalizing usage of new features.
Expand All @@ -41,17 +43,17 @@ public static void Example()
// Train the model.
var model = pipeline.Fit(trainingData);

// Create testing examples. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(500, seed:123));
// Create testing data. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(5, seed: 123));

// Run the model on test data set.
var transformedTestData = model.Transform(testData);

// Convert IDataView object to a list.
var predictions = mlContext.Data.CreateEnumerable<Prediction>(transformedTestData, reuseRowObject: false).ToList();

// Look at 5 predictions
foreach (var p in predictions.Take(5))
// Look at 5 predictions for the Label, side by side with the actual Label for comparison.
foreach (var p in predictions)
Console.WriteLine($"Label: {p.Label:F3}, Prediction: {p.Score:F3}");

// Expected output:
Expand All @@ -63,27 +65,26 @@ public static void Example()

// Evaluate the overall metrics
var metrics = mlContext.Regression.Evaluate(transformedTestData);
Microsoft.ML.SamplesUtils.ConsoleUtils.PrintMetrics(metrics);
PrintMetrics(metrics);

// Expected output:
// Mean Absolute Error: 0.06
// Mean Squared Error: 0.01
// Root Mean Squared Error: 0.08
// RSquared: 0.93
// Root Mean Squared Error: 0.07
// RSquared: 0.95 (closer to 1 is better. The worest case is 0)
}

private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed=0)
{
var random = new Random(seed);
float randomFloat() => (float)random.NextDouble();
for (int i = 0; i < count; i++)
{
var label = randomFloat();
float label = (float)random.NextDouble();
yield return new DataPoint
{
Label = label,
// Create random features that are correlated with label.
Features = Enumerable.Repeat(label, 50).Select(x => x + randomFloat()).ToArray()
// Create random features that are correlated with the label.
Features = Enumerable.Repeat(label, 50).Select(x => x + (float)random.NextDouble()).ToArray()
};
}
}
Expand All @@ -104,5 +105,15 @@ private class Prediction
// Predicted score from the trainer.
public float Score { get; set; }
}

// Print some evaluation metrics to regression problems.
private static void PrintMetrics(RegressionMetrics metrics)
{
Console.WriteLine($"Mean Absolute Error: {metrics.MeanAbsoluteError:F2}");
Console.WriteLine($"Mean Squared Error: {metrics.MeanSquaredError:F2}");
Console.WriteLine($"Root Mean Squared Error: {metrics.RootMeanSquaredError:F2}");
Console.WriteLine($"RSquared: {metrics.RSquared:F2}");
}
}
}
}

34 changes: 34 additions & 0 deletions docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Regression/FastForestWithOptions.tt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,34 @@
<#@ include file="RegressionSamplesTemplate.ttinclude"#>

<#+
string ClassHeader = @"// This example requires installation of additional NuGet package
// <a href=""https://www.nuget.org/packages/Microsoft.ML.FastTree/"">Microsoft.ML.FastTree</a>. ";

string ClassName="FastForestWithOptionsRegression";
string ExtraUsing = "using Microsoft.ML.Trainers.FastTree;";
string Trainer = @"FastForest";
string TrainerOptions = @"FastForestRegressionTrainer.Options
{
LabelColumnName = nameof(DataPoint.Label),
FeatureColumnName = nameof(DataPoint.Features),
// Only use 80% of features to reduce over-fitting.
FeatureFraction = 0.8,
// Create a simpler model by penalizing usage of new features.
Copy link
Member

@sfilipi sfilipi Apr 15, 2019 •
edited by wschin
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

reate a simpler model by penalizing usage of new features. [](start = 20, length = 58)

great idea, on putting on comments. #Resolved

FeatureFirstUsePenalty = 0.1,
// Reduce the number of trees to 50.
NumberOfTrees = 50
}";

string ExpectedOutputPerInstance= @"// Expected output:
// Label: 0.985, Prediction: 0.866
// Label: 0.155, Prediction: 0.171
// Label: 0.515, Prediction: 0.470
// Label: 0.566, Prediction: 0.476
// Label: 0.096, Prediction: 0.140";

string ExpectedOutput = @"// Expected output:
// Mean Absolute Error: 0.06
// Mean Squared Error: 0.01
// Root Mean Squared Error: 0.07
// RSquared: 0.95 (closer to 1 is better. The worest case is 0)";
#>
49 changes: 29 additions & 20 deletions docs/samples/Microsoft.ML.Samples/Dynamic/Trainers/Regression/FastTree.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,40 +6,40 @@

namespace Samples.Dynamic.Trainers.Regression
{
public static class FastTree
public static class FastTreeRegression
{
// This example requires installation of additional NuGet package
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
// <a href="https://www.nuget.org/packages/Microsoft.ML.FastTree/">Microsoft.ML.FastTree</a>.
public static void Example()
{
// Create a new context for ML.NET operations. It can be used for exception tracking and logging,
// as a catalog of available operations and as the source of randomness.
// Setting the seed to a fixed number in this example to make outputs deterministic.
var mlContext = new MLContext(seed: 0);

// Create a list of training examples.
var examples = GenerateRandomDataPoints(1000);
// Create a list of training data points.
var dataPoints = GenerateRandomDataPoints(1000);

// Convert the examples list to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(examples);
// Convert the list of data points to an IDataView object, which is consumable by ML.NET API.
var trainingData = mlContext.Data.LoadFromEnumerable(dataPoints);

// Define the trainer.
var pipeline = mlContext.Regression.Trainers.FastTree();
var pipeline = mlContext.Regression.Trainers.FastTree(labelColumnName: nameof(DataPoint.Label), featureColumnName: nameof(DataPoint.Features));

// Train the model.
var model = pipeline.Fit(trainingData);

// Create testing examples. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(500, seed:123));
// Create testing data. Use different random seed to make it different from training data.
var testData = mlContext.Data.LoadFromEnumerable(GenerateRandomDataPoints(5, seed: 123));

// Run the model on test data set.
var transformedTestData = model.Transform(testData);

// Convert IDataView object to a list.
var predictions = mlContext.Data.CreateEnumerable<Prediction>(transformedTestData, reuseRowObject: false).ToList();

// Look at 5 predictions
foreach (var p in predictions.Take(5))
// Look at 5 predictions for the Label, side by side with the actual Label for comparison.
foreach (var p in predictions)
Console.WriteLine($"Label: {p.Label:F3}, Prediction: {p.Score:F3}");

// Expected output:
Expand All @@ -51,27 +51,26 @@ public static void Example()

// Evaluate the overall metrics
var metrics = mlContext.Regression.Evaluate(transformedTestData);
Microsoft.ML.SamplesUtils.ConsoleUtils.PrintMetrics(metrics);
PrintMetrics(metrics);

// Expected output:
// Mean Absolute Error: 0.05
// Mean Absolute Error: 0.03
// Mean Squared Error: 0.00
// Root Mean Squared Error: 0.06
// RSquared: 0.95
// Root Mean Squared Error: 0.03
// RSquared: 0.99 (closer to 1 is better. The worest case is 0)
}

private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count, int seed=0)
{
var random = new Random(seed);
float randomFloat() => (float)random.NextDouble();
for (int i = 0; i < count; i++)
{
var label = randomFloat();
float label = (float)random.NextDouble();
yield return new DataPoint
{
Label = label,
// Create random features that are correlated with label.
Features = Enumerable.Repeat(label, 50).Select(x => x + randomFloat()).ToArray()
// Create random features that are correlated with the label.
Features = Enumerable.Repeat(label, 50).Select(x => x + (float)random.NextDouble()).ToArray()
};
}
}
Expand All @@ -92,5 +91,15 @@ private class Prediction
// Predicted score from the trainer.
public float Score { get; set; }
}

// Print some evaluation metrics to regression problems.
private static void PrintMetrics(RegressionMetrics metrics)
{
Console.WriteLine($"Mean Absolute Error: {metrics.MeanAbsoluteError:F2}");
Console.WriteLine($"Mean Squared Error: {metrics.MeanSquaredError:F2}");
Console.WriteLine($"Root Mean Squared Error: {metrics.RootMeanSquaredError:F2}");
Console.WriteLine($"RSquared: {metrics.RSquared:F2}");
}
}
}
}

Loading