static void Main(string[] args)
{
var assetsPath = ModelHelpers.GetAssetsPath(@"..\..\..\assets");
var transactionsCsv = Path.Combine(assetsPath, "inputs", "transactions.csv");
var offersCsv = Path.Combine(assetsPath, "inputs", "offers.csv");
var pivotCsv = Path.Combine(assetsPath, "inputs", "pivot.csv");
var modelZip = Path.Combine(assetsPath, "outputs", "retailClustering.zip");
try
{
//STEP 0: Special data pre-process in this sample creating the PivotTable csv file
DataHelpers.PreProcessAndSave(offersCsv, transactionsCsv, pivotCsv);
//Create the MLContext to share across components for deterministic results
MLContext mlContext = new MLContext(seed: 1); //Seed set to any number so you have a deterministic environment
// STEP 1: Common data loading configuration
var textLoader = CustomerSegmentationTextLoaderFactory.CreateTextLoader(mlContext);
var pivotDataView = textLoader.Read(pivotCsv);
//STEP 2: Configure data transformations in pipeline
var dataProcessPipeline = new PrincipalComponentAnalysisEstimator(mlContext, "Features", "PCAFeatures", rank: 2)
.Append(new OneHotEncodingEstimator(mlContext, new[] { new OneHotEncodingEstimator.ColumnInfo("LastName",
"LastNameKey",
CategoricalTransform.OutputKind.Ind) }));
// (Optional) Peek data in training DataView after applying the ProcessPipeline's transformations
Common.ConsoleHelper.PeekDataViewInConsole <PivotObservation>(mlContext, pivotDataView, dataProcessPipeline, 10);
Common.ConsoleHelper.PeekVectorColumnDataInConsole(mlContext, "Features", pivotDataView, dataProcessPipeline, 10);
// STEP 3: Create and train the model
var trainer = mlContext.Clustering.Trainers.KMeans("Features", clustersCount: 3);
var modelBuilder = new Common.ModelBuilder <PivotObservation, ClusteringPrediction>(mlContext, dataProcessPipeline);
modelBuilder.AddTrainer(trainer);
var trainedModel = modelBuilder.Train(pivotDataView);
// STEP4: Evaluate accuracy of the model
var metrics = modelBuilder.EvaluateClusteringModel(pivotDataView);
Common.ConsoleHelper.PrintClusteringMetrics(trainer.ToString(), metrics);
// STEP5: Save/persist the model as a .ZIP file
modelBuilder.SaveModelAsFile(modelZip);
} catch (Exception ex)
{
Common.ConsoleHelper.ConsoleWriteException(ex.Message);
}
Common.ConsoleHelper.ConsolePressAnyKey();
}
public void TestPcaEstimator()
{
var data = TextLoader.CreateReader(_env,
c => (label: c.LoadFloat(11), features: c.LoadFloat(0, 10)),
separator: ';', hasHeader: true)
.Read(_dataSource);
var est = new PrincipalComponentAnalysisEstimator(_env, "features", "pca", rank: 5, seed: 1);
var outputPath = GetOutputPath("PCA", "pca.tsv");
using (var ch = _env.Start("save"))
{
IDataView savedData = TakeFilter.Create(_env, est.Fit(data.AsDynamic).Transform(data.AsDynamic), 4);
savedData = ColumnSelectingTransformer.CreateKeep(_env, savedData, new[] { "pca" });
using (var fs = File.Create(outputPath))
DataSaverUtils.SaveDataView(ch, _saver, savedData, fs, keepHidden: true);
}
CheckEquality("PCA", "pca.tsv", digitsOfPrecision: 4);
Done();
}
public void PcaWorkout()
{
var data = TextLoader.CreateReader(_env,
c => (label: c.LoadFloat(11), weight: c.LoadFloat(0), features: c.LoadFloat(1, 10)),
separator: ';', hasHeader: true)
.Read(_dataSource);
var invalidData = TextLoader.CreateReader(_env,
c => (label: c.LoadFloat(11), weight: c.LoadFloat(0), features: c.LoadText(1, 10)),
separator: ';', hasHeader: true)
.Read(_dataSource);
var est = new PrincipalComponentAnalysisEstimator(_env, "features", "pca", rank: 4, seed: 10);
TestEstimatorCore(est, data.AsDynamic, invalidInput: invalidData.AsDynamic);
var estNonDefaultArgs = new PrincipalComponentAnalysisEstimator(_env, "features", "pca", rank: 3, weightColumn: "weight", overSampling: 2, center: false);
TestEstimatorCore(estNonDefaultArgs, data.AsDynamic, invalidInput: invalidData.AsDynamic);
Done();
}
static void Main(string[] args)
{
string assetsRelativePath = @"../../../assets";
string assetsPath = GetDataSetAbsolutePath(assetsRelativePath);
string transactionsCsv = Path.Combine(assetsPath, "inputs", "transactions.csv");
string offersCsv = Path.Combine(assetsPath, "inputs", "offers.csv");
string pivotCsv = Path.Combine(assetsPath, "inputs", "pivot.csv");
string modelZip = Path.Combine(assetsPath, "outputs", "retailClustering.zip");
try
{
//STEP 0: Special data pre-process in this sample creating the PivotTable csv file
DataHelpers.PreProcessAndSave(offersCsv, transactionsCsv, pivotCsv);
//Create the MLContext to share across components for deterministic results
MLContext mlContext = new MLContext(seed: 1); //Seed set to any number so you have a deterministic environment
// STEP 1: Common data loading configuration
var pivotDataView = mlContext.Data.ReadFromTextFile(path: pivotCsv,
columns: new[]
{
new TextLoader.Column(DefaultColumnNames.Features, DataKind.R4, new[] { new TextLoader.Range(0, 31) }),
new TextLoader.Column(nameof(PivotData.LastName), DataKind.Text, 32)
},
hasHeader: true,
separatorChar: ',');
//STEP 2: Configure data transformations in pipeline
var dataProcessPipeline = new PrincipalComponentAnalysisEstimator(env: mlContext, outputColumnName: "PCAFeatures", inputColumnName: DefaultColumnNames.Features, rank: 2)
.Append(new OneHotEncodingEstimator(mlContext,
new[]
{
new OneHotEncodingEstimator.ColumnInfo(name: "LastNameKey", inputColumnName: nameof(PivotData.LastName),
OneHotEncodingTransformer.OutputKind.Ind)
}
));
// (Optional) Peek data in training DataView after applying the ProcessPipeline's transformations
Common.ConsoleHelper.PeekDataViewInConsole(mlContext, pivotDataView, dataProcessPipeline, 10);
Common.ConsoleHelper.PeekVectorColumnDataInConsole(mlContext, DefaultColumnNames.Features, pivotDataView, dataProcessPipeline, 10);
//STEP 3: Create the training pipeline
var trainer = mlContext.Clustering.Trainers.KMeans(featureColumn: DefaultColumnNames.Features, clustersCount: 3);
var trainingPipeline = dataProcessPipeline.Append(trainer);
//STEP 4: Train the model fitting to the pivotDataView
Console.WriteLine("=============== Training the model ===============");
ITransformer trainedModel = trainingPipeline.Fit(pivotDataView);
//STEP 5: Evaluate the model and show accuracy stats
Console.WriteLine("===== Evaluating Model's accuracy with Test data =====");
var predictions = trainedModel.Transform(pivotDataView);
var metrics = mlContext.Clustering.Evaluate(predictions, score: DefaultColumnNames.Score, features: DefaultColumnNames.Features);
ConsoleHelper.PrintClusteringMetrics(trainer.ToString(), metrics);
//STEP 6: Save/persist the trained model to a .ZIP file
using (var fs = new FileStream(modelZip, FileMode.Create, FileAccess.Write, FileShare.Write))
mlContext.Model.Save(trainedModel, fs);
Console.WriteLine("The model is saved to {0}", modelZip);
}
catch (Exception ex)
{
Common.ConsoleHelper.ConsoleWriteException(ex.Message);
}
Common.ConsoleHelper.ConsolePressAnyKey();
}
static void Main(string[] args)
{
var assetsPath = PathHelper.GetAssetsPath(@"..\..\..\assets");
var transactionsCsv = Path.Combine(assetsPath, "inputs", "transactions.csv");
var offersCsv = Path.Combine(assetsPath, "inputs", "offers.csv");
var pivotCsv = Path.Combine(assetsPath, "inputs", "pivot.csv");
var modelZip = Path.Combine(assetsPath, "outputs", "retailClustering.zip");
try
{
//STEP 0: Special data pre-process in this sample creating the PivotTable csv file
DataHelpers.PreProcessAndSave(offersCsv, transactionsCsv, pivotCsv);
//Create the MLContext to share across components for deterministic results
MLContext mlContext = new MLContext(seed: 1); //Seed set to any number so you have a deterministic environment
// STEP 1: Common data loading configuration
TextLoader textLoader = mlContext.Data.TextReader(new TextLoader.Arguments()
{
Separator = ",",
HasHeader = true,
Column = new[]
{
new TextLoader.Column("Features", DataKind.R4, new[] { new TextLoader.Range(0, 31) }),
new TextLoader.Column("LastName", DataKind.Text, 32)
}
});
var pivotDataView = textLoader.Read(pivotCsv);
//STEP 2: Configure data transformations in pipeline
var dataProcessPipeline = new PrincipalComponentAnalysisEstimator(mlContext, "Features", "PCAFeatures", rank: 2)
.Append(new OneHotEncodingEstimator(mlContext, new[] { new OneHotEncodingEstimator.ColumnInfo("LastName",
"LastNameKey",
CategoricalTransform.OutputKind.Ind) }));
// (Optional) Peek data in training DataView after applying the ProcessPipeline's transformations
Common.ConsoleHelper.PeekDataViewInConsole <PivotObservation>(mlContext, pivotDataView, dataProcessPipeline, 10);
Common.ConsoleHelper.PeekVectorColumnDataInConsole(mlContext, "Features", pivotDataView, dataProcessPipeline, 10);
//STEP 3: Create the training pipeline
var trainer = mlContext.Clustering.Trainers.KMeans("Features", clustersCount: 3);
var trainingPipeline = dataProcessPipeline.Append(trainer);
//STEP 4: Train the model fitting to the pivotDataView
Console.WriteLine("=============== Training the model ===============");
ITransformer trainedModel = trainingPipeline.Fit(pivotDataView);
//STEP 5: Evaluate the model and show accuracy stats
Console.WriteLine("===== Evaluating Model's accuracy with Test data =====");
var predictions = trainedModel.Transform(pivotDataView);
var metrics = mlContext.Clustering.Evaluate(predictions, score: "Score", features: "Features");
ConsoleHelper.PrintClusteringMetrics(trainer.ToString(), metrics);
//STEP 6: Save/persist the trained model to a .ZIP file
using (var fs = new FileStream(modelZip, FileMode.Create, FileAccess.Write, FileShare.Write))
mlContext.Model.Save(trainedModel, fs);
Console.WriteLine("The model is saved to {0}", modelZip);
}
catch (Exception ex)
{
Common.ConsoleHelper.ConsoleWriteException(ex.Message);
}
Common.ConsoleHelper.ConsolePressAnyKey();
}
