exception fixes (dotnet#136)

src/Microsoft.ML.Auto/ColumnInference/ColumnInferenceApi.cs

@@ -2,7 +2,6 @@
 // The .NET Foundation licenses this file to you under the MIT license.
 // See the LICENSE file in the project root for more information.
 
-using System;
 using System.Collections.Generic;
 using System.Linq;
 using Microsoft.ML.Data;
@@ -16,13 +15,7 @@ public static (TextLoader.Arguments, IEnumerable<(string, ColumnPurpose)>) Infer
         {
             var sample = TextFileSample.CreateFromFullFile(path);
             var splitInference = InferSplit(sample, separatorChar, allowQuotedStrings, supportSparse);
-            var typeInference = InferColumnTypes(context, sample, splitInference, hasHeader);
-
-            // If label column index > inferred # of columns, throw error
-            if (labelColumnIndex >= typeInference.Columns.Count())
-            {
-                throw new ArgumentOutOfRangeException(nameof(labelColumnIndex), $"Label column index ({labelColumnIndex}) is >= than # of inferred columns ({typeInference.Columns.Count()}).");
-            }
+            var typeInference = InferColumnTypes(context, sample, splitInference, hasHeader, labelColumnIndex, null);
 
             // if no column is named label,
             // rename label column to default ML.NET label column name
@@ -40,7 +33,7 @@ public static (TextLoader.Arguments, IEnumerable<(string, ColumnPurpose)>) Infer
         {
             var sample = TextFileSample.CreateFromFullFile(path);
             var splitInference = InferSplit(sample, separatorChar, allowQuotedStrings, supportSparse);
-            var typeInference = InferColumnTypes(context, sample, splitInference, true);
+            var typeInference = InferColumnTypes(context, sample, splitInference, true, null, label);
             return InferColumns(context, path, label, true, splitInference, typeInference, trimWhitespace, groupColumns);
         }
 
@@ -49,10 +42,6 @@ public static (TextLoader.Arguments, IEnumerable<(string, ColumnPurpose)>) Infer
             bool trimWhitespace, bool groupColumns)
         {
             var loaderColumns = ColumnTypeInference.GenerateLoaderColumns(typeInference.Columns);
-            if (!loaderColumns.Any(t => label.Equals(t.Name)))
-            {
-                throw new InferenceException(InferenceType.Label, $"Specified Label Column '{label}' was not found.");
-            }
             var typedLoaderArgs = new TextLoader.Arguments
             {
                 Column = loaderColumns,
@@ -121,7 +110,7 @@ private static TextFileContents.ColumnSplitResult InferSplit(TextFileSample samp
         }
 
         private static ColumnTypeInference.InferenceResult InferColumnTypes(MLContext context, TextFileSample sample,
-            TextFileContents.ColumnSplitResult splitInference, bool hasHeader)
+            TextFileContents.ColumnSplitResult splitInference, bool hasHeader, uint? labelColumnIndex, string label)
         {
             // infer column types
             var typeInferenceResult = ColumnTypeInference.InferTextFileColumnTypes(context, sample,
@@ -131,7 +120,9 @@ private static ColumnTypeInference.InferenceResult InferColumnTypes(MLContext co
                     Separator = splitInference.Separator.Value,
                     AllowSparse = splitInference.AllowSparse,
                     AllowQuote = splitInference.AllowQuote,
-                    HasHeader = hasHeader
+                    HasHeader = hasHeader,
+                    LabelColumnIndex = labelColumnIndex,
+                    Label = label
                 });
 
             if (!typeInferenceResult.IsSuccess)

src/Microsoft.ML.Auto/ColumnInference/ColumnTypeInference.cs

@@ -30,6 +30,8 @@ internal sealed class Arguments
             public int ColumnCount;
             public bool HasHeader;
             public int MaxRowsToRead;
+            public uint? LabelColumnIndex;
+            public string Label;
 
             public Arguments()
             {
@@ -68,13 +70,31 @@ public IntermediateColumn(ReadOnlyMemory<char>[] data, int columnId)
             }
 
             public ReadOnlyMemory<char>[] RawData { get { return _data; } }
+
+            public string Name { get; set; }
+
+            public bool HasAllBooleanValues()
+            {
+                if (this.RawData.Skip(1)
+                    .All(x => {
+                        bool value;
+                        // (note: Conversions.TryParse parses an empty string as a Boolean)
+                        return !string.IsNullOrEmpty(x.ToString()) &&
+                            Conversions.TryParse(in x, out value);
+                    }))
+                {
+                    return true;
+                }
+
+                return false;
+            }
         }
 
-        public struct Column
+        public class Column
         {
             public readonly int ColumnIndex;
-            public readonly PrimitiveType ItemType;
 
+            public PrimitiveType ItemType;
             public string SuggestedName;
 
             public Column(int columnIndex, string suggestedName, PrimitiveType itemType)
@@ -131,13 +151,10 @@ public void Apply(IntermediateColumn[] columns)
                 {
                     foreach (var col in columns)
                     {
-                        if (!col.RawData.Skip(1)
-                            .All(x =>
-                            {
-                                bool value;
-                                return Conversions.TryParse(in x, out value);
-                            })
-                            )
+                        // skip columns that already have a suggested type,
+                        // or that don't have all Boolean values
+                        if (col.SuggestedType != null ||
+                            !col.HasAllBooleanValues())
                         {
                             continue;
                         }
@@ -156,12 +173,6 @@ public void Apply(IntermediateColumn[] columns)
                 {
                     foreach (var col in columns)
                     {
-                        // skip columns that already have a suggested type
-                        if(col.SuggestedType != null)
-                        {
-                            continue;
-                        }
-
                         if (!col.RawData.Skip(1)
                             .All(x =>
                             {
@@ -215,9 +226,9 @@ public void Apply(IntermediateColumn[] columns)
         private static IEnumerable<ITypeInferenceExpert> GetExperts()
         {
             // Current logic is pretty primitive: if every value (except the first) of a column
-            // parses as a boolean it's boolean, if it parses as numeric then it's numeric. Otherwise, it is text.
-            yield return new Experts.BooleanValues();
+            // parses as numeric then it's numeric. Else if it parses as a Boolean, it's Boolean. Otherwise, it is text.
             yield return new Experts.AllNumericValues();
+            yield return new Experts.BooleanValues();
             yield return new Experts.EverythingText();
         }
 
@@ -329,7 +340,6 @@ private static InferenceResult InferTextFileColumnTypesCore(MLContext env, IMult
             }
 
             // suggest names
-            var names = new List<string>();
             usedNames.Clear();
             foreach (var col in cols)
             {
@@ -338,14 +348,23 @@ private static InferenceResult InferTextFileColumnTypesCore(MLContext env, IMult
                 name0 = name = SuggestName(col, args.HasHeader);
                 int i = 0;
                 while (!usedNames.Add(name))
+                {
                     name = string.Format("{0}_{1:00}", name0, i++);
-                names.Add(name);
+                }
+                col.Name = name;
+            }
+
+            // validate & retrieve label column
+            var labelColumn = GetAndValidateLabelColumn(args, cols);
+
+            // if label column has all Boolean values, set its type as Boolean
+            if(labelColumn.HasAllBooleanValues())
+            {
+                labelColumn.SuggestedType = BoolType.Instance;
             }
-            var outCols =
-                cols.Select((x, i) => new Column(x.ColumnId, names[i], x.SuggestedType)).ToArray();
 
-            var numerics = outCols.Count(x => x.ItemType.IsNumber());
-            
+            var outCols = cols.Select(x => new Column(x.ColumnId, x.Name, x.SuggestedType)).ToArray();
+
             return InferenceResult.Success(outCols, args.HasHeader, cols.Select(col => col.RawData).ToArray());
         }
 
@@ -361,6 +380,31 @@ private static string Sanitize(string header)
             return string.Join("", header.Select(x => Char.IsLetterOrDigit(x) ? x : '_'));
         }
 
+        private static IntermediateColumn GetAndValidateLabelColumn(Arguments args, IntermediateColumn[] cols)
+        {
+            IntermediateColumn labelColumn = null;
+            if (args.LabelColumnIndex != null)
+            {
+                // if label column index > inferred # of columns, throw error
+                if (args.LabelColumnIndex >= cols.Count())
+                {
+                    throw new ArgumentOutOfRangeException(nameof(args.LabelColumnIndex), $"Label column index ({args.LabelColumnIndex}) is >= than # of inferred columns ({cols.Count()}).");
+                }
+
+                labelColumn = cols[args.LabelColumnIndex.Value];
+            }
+            else
+            {
+                labelColumn = cols.FirstOrDefault(c => c.Name == args.Label);
+                if (labelColumn == null)
+                {
+                    throw new ArgumentException($"Specified label column '{args.Label}' was not found.");
+                }
+            }
+
+            return labelColumn;
+        }
+
         public static TextLoader.Column[] GenerateLoaderColumns(Column[] columns)
         {
             var loaderColumns = new List<TextLoader.Column>();

src/Microsoft.ML.Auto/ColumnInference/PurposeInference.cs

@@ -171,7 +171,9 @@ public void Apply(IntermediateColumn[] columns)
                             Double avgSpaces = 1.0 * sumSpaces / data.Length;
                             if (cardinalityRatio < 0.7 || seen.Count < 100)
                                 column.SuggestedPurpose = ColumnPurpose.CategoricalFeature;
-                            else if (cardinalityRatio >= 0.85 && (avgLength > 30 || avgSpaces >= 1))
+                            // (note: the columns.Count() == 1 condition below, in case a dataset has only
+                            // a 'name' and a 'label' column, forces what would be a 'name' column to become a text feature)
+                            else if (cardinalityRatio >= 0.85 && (avgLength > 30 || avgSpaces >= 1 || columns.Count() == 1))
                                 column.SuggestedPurpose = ColumnPurpose.TextFeature;
                             else if (cardinalityRatio >= 0.9)
                                 column.SuggestedPurpose = ColumnPurpose.Name;

src/Microsoft.ML.Auto/Utils/UserInputValidationUtil.cs

@@ -34,18 +34,13 @@ public static void ValidateInferColumnsArgs(string path)
             ValidatePath(path);
         }
 
-        public static void ValidateAutoReadArgs(string path, string label)
-        {
-            ValidateLabel(label);
-            ValidatePath(path);
-        }
-
         private static void ValidateTrainData(IDataView trainData)
         {
             if (trainData == null)
             {
                 throw new ArgumentNullException(nameof(trainData), "Training data cannot be null");
             }
+
             var type = trainData.Schema.GetColumnOrNull(DefaultColumnNames.Features)?.Type.GetItemType();
             if (type != null && type != NumberType.R4)
             {

src/Test/ColumnInferenceTests.cs

@@ -28,7 +28,7 @@ public void IncorrectLabelColumnTest()
         {
             var dataPath = DatasetUtil.DownloadUciAdultDataset();
             var context = new MLContext();
-            Assert.ThrowsException<InferenceException>(new System.Action(() => context.Data.InferColumns(dataPath, "Junk", groupColumns: false)));
+            Assert.ThrowsException<ArgumentException>(new System.Action(() => context.Data.InferColumns(dataPath, "Junk", groupColumns: false)));
         }
 
         [TestMethod]
@@ -62,5 +62,51 @@ public void InferColumnsLabelIndexNoHeaders()
             Assert.AreEqual(1, labelPurposes.Count());
             Assert.AreEqual(DefaultColumnNames.Label, labelPurposes.First().Name);
         }
+
+        [TestMethod]
+        public void InferColumnsWithDatasetWithEmptyColumn()
+        {
+            var result = new MLContext().Data.InferColumns(@".\TestData\DatasetWithEmptyColumn.txt", DefaultColumnNames.Label);
+            var emptyColumn = result.TextLoaderArgs.Column.First(c => c.Name == "Empty");
+            Assert.AreEqual(DataKind.TX, emptyColumn.Type);
+        }
+
+        [TestMethod]
+        public void InferColumnsWithDatasetWithBoolColumn()
+        {
+            var result = new MLContext().Data.InferColumns(@".\TestData\BinaryDatasetWithBoolColumn.txt", DefaultColumnNames.Label);
+            Assert.AreEqual(2, result.TextLoaderArgs.Column.Count());
+            Assert.AreEqual(2, result.ColumnPurpopses.Count());
+
+            var boolColumn = result.TextLoaderArgs.Column.First(c => c.Name == "Bool");
+            var labelColumn = result.TextLoaderArgs.Column.First(c => c.Name == DefaultColumnNames.Label);
+            // ensure non-label Boolean column is detected as R4
+            Assert.AreEqual(DataKind.R4, boolColumn.Type);
+            Assert.AreEqual(DataKind.BL, labelColumn.Type);
+
+            // ensure non-label Boolean column is detected as R4
+            var boolPurpose = result.ColumnPurpopses.First(c => c.Name == "Bool").Purpose;
+            var labelPurpose = result.ColumnPurpopses.First(c => c.Name == DefaultColumnNames.Label).Purpose;
+            Assert.AreEqual(ColumnPurpose.NumericFeature, boolPurpose);
+            Assert.AreEqual(ColumnPurpose.Label, labelPurpose);
+        }
+
+        [TestMethod]
+        public void InferColumnsWhereNameColumnIsOnlyFeature()
+        {
+            var result = new MLContext().Data.InferColumns(@".\TestData\NameColumnIsOnlyFeatureDataset.txt", DefaultColumnNames.Label);
+            Assert.AreEqual(2, result.TextLoaderArgs.Column.Count());
+            Assert.AreEqual(2, result.ColumnPurpopses.Count());
+
+            var nameColumn = result.TextLoaderArgs.Column.First(c => c.Name == DefaultColumnNames.Name);
+            var labelColumn = result.TextLoaderArgs.Column.First(c => c.Name == DefaultColumnNames.Label);
+            Assert.AreEqual(DataKind.TX, nameColumn.Type);
+            Assert.AreEqual(DataKind.BL, labelColumn.Type);
+
+            var namePurpose = result.ColumnPurpopses.First(c => c.Name == DefaultColumnNames.Name).Purpose;
+            var labelPurpose = result.ColumnPurpopses.First(c => c.Name == DefaultColumnNames.Label).Purpose;
+            Assert.AreEqual(ColumnPurpose.TextFeature, namePurpose);
+            Assert.AreEqual(ColumnPurpose.Label, labelPurpose);
+        }
     }
 }

src/Test/Test.csproj

@@ -18,4 +18,16 @@
     <ProjectReference Include="..\Microsoft.ML.Auto\Microsoft.ML.Auto.csproj" />
   </ItemGroup>
 
+  <ItemGroup>
+    <None Update="TestData\NameColumnIsOnlyFeatureDataset.txt">
+      <CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
+    </None>
+    <None Update="TestData\BinaryDatasetWithBoolColumn.txt">
+      <CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
+    </None>
+    <None Update="TestData\DatasetWithEmptyColumn.txt">
+      <CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
+    </None>
+  </ItemGroup>
+
 </Project>

src/Test/TestData/BinaryDatasetWithBoolColumn.txt

@@ -0,0 +1,5 @@
+Label,Bool
+0,1
+0,0
+1,1
+1,0

src/Test/TestData/DatasetWithEmptyColumn.txt

@@ -0,0 +1,4 @@
+Label,Feature1,Empty
+0,2,
+0,4,
+1,1,