Skip to content

Fix for KeytoValue transformer #4866

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 3 commits into from
Feb 21, 2020
Merged

Fix for KeytoValue transformer #4866

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
73d3318
fix for keytovalue
Lynx1820 Feb 20, 2020
216ddba
fix for keytovalue
Lynx1820 Feb 20, 2020
466120f
resolving comments
Lynx1820 Feb 20, 2020
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
4 changes: 2 additions & 2 deletions src/Microsoft.ML.Data/Transforms/KeyToValue.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -568,9 +568,9 @@ public void SaveAsOnnx(OnnxContext ctx)

if (!ctx.ContainsColumn(inputColumnName))
continue;

string srcVariableName = ctx.GetVariableName(inputColumnName);
var dstVariableName = ctx.AddIntermediateVariable(_types[iinfo], info.outputColumnName, true);
if (!_kvMaps[iinfo].SaveOnnx(ctx, inputColumnName, dstVariableName))
if (!_kvMaps[iinfo].SaveOnnx(ctx, srcVariableName, dstVariableName))
Copy link
Contributor

Choose a reason for hiding this comment

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

I think it would be a good idea to add a test case specifically because we don't have tests for this. Also please verify this fixes the Nimbus issue.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Sounds good. I restructured the tests and got rid of a KeyToValue test that seems redundant and is not formatted to test different types of values.

{
ctx.RemoveColumn(inputColumnName, true);
}
Expand Down
86 changes: 47 additions & 39 deletions test/Microsoft.ML.Tests/OnnxConversionTest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@
using System.Runtime.InteropServices;
using System.Text.RegularExpressions;
using Google.Protobuf;
using Google.Protobuf.WellKnownTypes;
using Microsoft.ML.Data;
using Microsoft.ML.EntryPoints;
using Microsoft.ML.Model.OnnxConverter;
Expand Down Expand Up @@ -1150,7 +1151,7 @@ public void IndicateMissingValuesOnnxConversionTest()
[InlineData(DataKind.Int64)]
[InlineData(DataKind.Double)]
[InlineData(DataKind.String)]
public void ValueToKeyandKeyToValueMappingOnnxConversionTest(DataKind valueType)
public void ValueToKeyMappingOnnxConversionTest(DataKind valueType)
{
var mlContext = new MLContext(seed: 1);
string filePath = GetDataPath("type-conversion.txt");
Expand All @@ -1160,9 +1161,8 @@ public void ValueToKeyandKeyToValueMappingOnnxConversionTest(DataKind valueType)
new TextLoader.Column("Value", valueType, 0, 0)
};
var dataView = mlContext.Data.LoadFromTextFile(filePath, columns);
var pipeline = mlContext.Transforms.Conversion.MapValueToKey("Key", "Value");

var pipeline = mlContext.Transforms.Conversion.MapValueToKey("Key", "Value").
Append(mlContext.Transforms.Conversion.MapKeyToValue("ValueOutput", "Key"));
var model = pipeline.Fit(dataView);
var mlnetResult = model.Transform(dataView);

Expand All @@ -1176,13 +1176,55 @@ public void ValueToKeyandKeyToValueMappingOnnxConversionTest(DataKind valueType)
var onnxEstimator = mlContext.Transforms.ApplyOnnxModel(onnxModelPath);
var onnxTransformer = onnxEstimator.Fit(dataView);
var onnxResult = onnxTransformer.Transform(dataView);

CompareResults("ValueOutput", "ValueOutput", mlnetResult, onnxResult);
CompareSelectedColumns<uint>("Key", "Key", mlnetResult, onnxResult);
}
Done();
}

[Theory]
[InlineData(DataKind.Single)]
[InlineData(DataKind.Int64)]
[InlineData(DataKind.Double)]
[InlineData(DataKind.String)]
public void KeyToValueMappingOnnxConversionTest(DataKind valueType)
{
var mlContext = new MLContext(seed: 1);
string filePath = GetDataPath("type-conversion.txt");

TextLoader.Column[] columns = new[]
{
new TextLoader.Column("Value", valueType, 0, 0)
};
var dataView = mlContext.Data.LoadFromTextFile(filePath, columns);
IEstimator<ITransformer>[] pipelines =
{
mlContext.Transforms.Conversion.MapValueToKey("Key", "Value").
Append(mlContext.Transforms.Conversion.MapKeyToValue("Value", "Key")),

mlContext.Transforms.Conversion.MapValueToKey("Value").
Append(mlContext.Transforms.Conversion.MapKeyToValue("Value"))
};
for (int i = 0; i < pipelines.Length; i++)
{
var model = pipelines[i].Fit(dataView);
var mlnetResult = model.Transform(dataView);

var onnxModel = mlContext.Model.ConvertToOnnxProtobuf(model, dataView);
var onnxFileName = "KeyToValue.onnx";
var onnxModelPath = GetOutputPath(onnxFileName);
SaveOnnxModel(onnxModel, onnxModelPath, null);

if (IsOnnxRuntimeSupported())
{
var onnxEstimator = mlContext.Transforms.ApplyOnnxModel(onnxModelPath);
var onnxTransformer = onnxEstimator.Fit(dataView);
var onnxResult = onnxTransformer.Transform(dataView);
CompareResults("Value", "Value", mlnetResult, onnxResult);
}
}
Done();
}

private class TextData
{
public string Text { get; set; }
Expand Down Expand Up @@ -1355,40 +1397,6 @@ public void OptionalColumnOnnxTest(DataKind dataKind)
Done();
}

[Fact]
public void KeyToValueOnnxConversionTest()
{
var mlContext = new MLContext(seed: 1);

string dataPath = GetDataPath("breast-cancer.txt");
var dataView = mlContext.Data.LoadFromTextFile<BreastCancerMulticlassExample>(dataPath,
separatorChar: '\t',
hasHeader: true);

var pipeline = mlContext.Transforms.Conversion.MapValueToKey("LabelKey", "Label").
Append(mlContext.Transforms.Conversion.MapKeyToValue("LabelValue", "LabelKey"));

var model = pipeline.Fit(dataView);
var transformedData = model.Transform(dataView);
var onnxModel = mlContext.Model.ConvertToOnnxProtobuf(model, dataView);

var onnxFileName = "KeyToValue.onnx";
var onnxModelPath = GetOutputPath(onnxFileName);

SaveOnnxModel(onnxModel, onnxModelPath, null);

if (IsOnnxRuntimeSupported())
{
// Evaluate the saved ONNX model using the data used to train the ML.NET pipeline.
var onnxEstimator = mlContext.Transforms.ApplyOnnxModel(onnxModelPath);
var onnxTransformer = onnxEstimator.Fit(dataView);
var onnxResult = onnxTransformer.Transform(dataView);
CompareSelectedColumns<ReadOnlyMemory<Char>>("LabelValue", "LabelValue", transformedData, onnxResult);
}

Done();
}

[Fact]
public void MulticlassTrainersOnnxConversionTest()
{
Expand Down