Making Schema implement ISchema explicitly (#1894)

* Removed ColumnCount and GetColumns
* Removed GetColumnName and GetColumnType
* Removed GetMetadataTypes and GetMetadataTypeOrNull
* Removed GetMetadata and hid TryGetColumnIndex

Author: Zruty0

docs/samples/Microsoft.ML.Samples/Dynamic/FeatureContributionCalculationTransform.cs

@@ -85,7 +85,7 @@ public static void FeatureContributionCalculationTransform_Regression()
                 var value = row.Features[featureOfInterest];
                 var contribution = row.FeatureContributions[featureOfInterest];
                 var percentContribution = 100 * contribution / row.Score;
-                var name = data.Schema.GetColumnName(featureOfInterest + 1);
+                var name = data.Schema[(int) (featureOfInterest + 1)].Name;
                 var weight = weights.GetValues()[featureOfInterest];
 
                 Console.WriteLine("{0:0.00}\t{1:0.00}\t{2}\t{3:0.00}\t{4:0.00}\t{5:0.00}\t{6:0.00}",

docs/samples/Microsoft.ML.Samples/Dynamic/GeneralizedAdditiveModels.cs

@@ -48,8 +48,8 @@ public static void RunExample()
             // and use a small number of bins to make it easy to visualize in the console window.
             // For real appplications, it is recommended to start with the default number of bins.
             var labelName = "MedianHomeValue";
-            var featureNames = data.Schema.GetColumns()
-                .Select(tuple => tuple.column.Name) // Get the column names
+            var featureNames = data.Schema
+                .Select(column => column.Name) // Get the column names
                 .Where(name => name != labelName) // Drop the Label
                 .ToArray();
             var pipeline = mlContext.Transforms.Concatenate("Features", featureNames)

docs/samples/Microsoft.ML.Samples/Dynamic/PermutationFeatureImportance.cs

@@ -71,8 +71,8 @@ public static void PFI_Regression()
             // Now let's look at which features are most important to the model overall
             // First, we have to prepare the data:
             // Get the feature names as an IEnumerable
-            var featureNames = data.Schema.GetColumns()
-                .Select(tuple => tuple.column.Name) // Get the column names
+            var featureNames = data.Schema
+                .Select(column => column.Name) // Get the column names
                 .Where(name => name != labelName) // Drop the Label
                 .ToArray();
 

src/Microsoft.ML.Core/Data/IEstimator.cs

@@ -177,15 +177,15 @@ internal static SchemaShape Create(Schema schema)
                 {
                     // First create the metadata.
                     var mCols = new List<Column>();
-                    foreach (var metaNameType in schema.GetMetadataTypes(iCol))
+                    foreach (var metaColumn in schema[iCol].Metadata.Schema)
                     {
-                        GetColumnTypeShape(metaNameType.Value, out var mVecKind, out var mItemType, out var mIsKey);
-                        mCols.Add(new Column(metaNameType.Key, mVecKind, mItemType, mIsKey));
+                        GetColumnTypeShape(metaColumn.Type, out var mVecKind, out var mItemType, out var mIsKey);
+                        mCols.Add(new Column(metaColumn.Name, mVecKind, mItemType, mIsKey));
                     }
                     var metadata = mCols.Count > 0 ? new SchemaShape(mCols) : _empty;
                     // Next create the single column.
-                    GetColumnTypeShape(schema.GetColumnType(iCol), out var vecKind, out var itemType, out var isKey);
-                    cols.Add(new Column(schema.GetColumnName(iCol), vecKind, itemType, isKey, metadata));
+                    GetColumnTypeShape(schema[iCol].Type, out var vecKind, out var itemType, out var isKey);
+                    cols.Add(new Column(schema[iCol].Name, vecKind, itemType, isKey, metadata));
                 }
             }
             return new SchemaShape(cols);

src/Microsoft.ML.Core/Data/ISchemaBindableMapper.cs

@@ -87,7 +87,7 @@ public interface IRowToRowMapper
 
         /// <summary>
         /// Given a predicate specifying which columns are needed, return a predicate indicating which input columns are
-        /// needed. The domain of the function is defined over the indices of the columns of <see cref="Schema.ColumnCount"/>
+        /// needed. The domain of the function is defined over the indices of the columns of <see cref="Schema.Count"/>
         /// for <see cref="InputSchema"/>.
         /// </summary>
         Func<int, bool> GetDependencies(Func<int, bool> predicate);

src/Microsoft.ML.Core/Data/MetadataUtils.cs

@@ -239,13 +239,13 @@ public static uint GetMaxMetadataKind(this Schema schema, out int colMax, string
             colMax = -1;
             for (int col = 0; col < schema.Count; col++)
             {
-                var columnType = schema.GetMetadataTypeOrNull(metadataKind, col);
+                var columnType = schema[col].Metadata.Schema.GetColumnOrNull(metadataKind)?.Type;
                 if (columnType == null || !columnType.IsKey || columnType.RawKind != DataKind.U4)
                     continue;
                 if (filterFunc != null && !filterFunc(schema, col))
                     continue;
                 uint value = 0;
-                schema.GetMetadata(metadataKind, col, ref value);
+                schema[col].Metadata.GetValue(metadataKind, ref value);
                 if (max < value)
                 {
                     max = value;
@@ -264,11 +264,11 @@ internal static IEnumerable<int> GetColumnSet(this Schema schema, string metadat
         {
             for (int col = 0; col < schema.Count; col++)
             {
-                var columnType = schema.GetMetadataTypeOrNull(metadataKind, col);
+                var columnType = schema[col].Metadata.Schema.GetColumnOrNull(metadataKind)?.Type;
                 if (columnType != null && columnType.IsKey && columnType.RawKind == DataKind.U4)
                 {
                     uint val = 0;
-                    schema.GetMetadata(metadataKind, col, ref val);
+                    schema[col].Metadata.GetValue(metadataKind, ref val);
                     if (val == value)
                         yield return col;
                 }
@@ -284,11 +284,11 @@ internal static IEnumerable<int> GetColumnSet(this Schema schema, string metadat
         {
             for (int col = 0; col < schema.Count; col++)
             {
-                var columnType = schema.GetMetadataTypeOrNull(metadataKind, col);
+                var columnType = schema[col].Metadata.Schema.GetColumnOrNull(metadataKind)?.Type;
                 if (columnType != null && columnType.IsText)
                 {
                     ReadOnlyMemory<char> val = default;
-                    schema.GetMetadata(metadataKind, col, ref val);
+                    schema[col].Metadata.GetValue(metadataKind, ref val);
                     if (ReadOnlyMemoryUtils.EqualsStr(value, val))
                         yield return col;
                 }
@@ -338,7 +338,7 @@ internal static void GetSlotNames(RoleMappedSchema schema, RoleMappedSchema.Colu
                 VBufferUtils.Resize(ref slotNames, vectorSize, 0);
             }
             else
-                schema.Schema.GetMetadata(Kinds.SlotNames, list[0].Index, ref slotNames);
+                schema.Schema[list[0].Index].Metadata.GetValue(Kinds.SlotNames, ref slotNames);
         }
 
         [BestFriend]
@@ -447,14 +447,14 @@ internal static bool TryGetCategoricalFeatureIndices(Schema schema, int colIndex
 
             bool isValid = false;
             categoricalFeatures = null;
-            if (!(schema.GetColumnType(colIndex) is VectorType vecType && vecType.Size > 0))
+            if (!(schema[colIndex].Type is VectorType vecType && vecType.Size > 0))
                 return isValid;
 
-            var type = schema.GetMetadataTypeOrNull(MetadataUtils.Kinds.CategoricalSlotRanges, colIndex);
+            var type = schema[colIndex].Metadata.Schema.GetColumnOrNull(MetadataUtils.Kinds.CategoricalSlotRanges)?.Type;
             if (type?.RawType == typeof(VBuffer<int>))
             {
                 VBuffer<int> catIndices = default(VBuffer<int>);
-                schema.GetMetadata(MetadataUtils.Kinds.CategoricalSlotRanges, colIndex, ref catIndices);
+                schema[colIndex].Metadata.GetValue(MetadataUtils.Kinds.CategoricalSlotRanges, ref catIndices);
                 VBufferUtils.Densify(ref catIndices);
                 int columnSlotsCount = vecType.Size;
                 if (catIndices.Length > 0 && catIndices.Length % 2 == 0 && catIndices.Length <= columnSlotsCount * 2)

src/Microsoft.ML.Core/Data/RoleMappedSchema.cs

@@ -53,7 +53,7 @@ public static bool TryCreateFromName(Schema schema, string name, out ColumnInfo
             if (!schema.TryGetColumnIndex(name, out int index))
                 return false;
 
-            colInfo = new ColumnInfo(name, index, schema.GetColumnType(index));
+            colInfo = new ColumnInfo(name, index, schema[index].Type);
             return true;
         }
 
@@ -65,9 +65,9 @@ public static bool TryCreateFromName(Schema schema, string name, out ColumnInfo
         public static ColumnInfo CreateFromIndex(Schema schema, int index)
         {
             Contracts.CheckValue(schema, nameof(schema));
-            Contracts.CheckParam(0 <= index && index < schema.ColumnCount, nameof(index));
+            Contracts.CheckParam(0 <= index && index < schema.Count, nameof(index));
 
-            return new ColumnInfo(schema.GetColumnName(index), index, schema.GetColumnType(index));
+            return new ColumnInfo(schema[index].Name, index, schema[index].Type);
         }
     }
 

src/Microsoft.ML.Core/Data/Schema.cs

@@ -27,11 +27,6 @@ public sealed class Schema : ISchema, IReadOnlyList<Schema.Column>
         private readonly Column[] _columns;
         private readonly Dictionary<string, int> _nameMap;
 
-        /// <summary>
-        /// Number of columns in the schema.
-        /// </summary>
-        public int ColumnCount => _columns.Length;
-
         /// <summary>
         /// Number of columns in the schema.
         /// </summary>
@@ -249,7 +244,7 @@ public void GetValue<TValue>(string kind, ref TValue value)
                 GetGetter<TValue>(column.Value.Index)(ref value);
             }
 
-            public override string ToString() => string.Join(", ", Schema.GetColumns().Select(x => x.column.Name));
+            public override string ToString() => string.Join(", ", Schema.Select(x => x.Name));
 
         }
 
@@ -270,11 +265,6 @@ internal Schema(Column[] columns)
             }
         }
 
-        /// <summary>
-        /// Get all non-hidden columns as pairs of (index, <see cref="Column"/>).
-        /// </summary>
-        public IEnumerable<(int index, Column column)> GetColumns() => _nameMap.Values.Select(idx => (idx, _columns[idx]));
-
         /// <summary>
         /// Manufacture an instance of <see cref="Schema"/> out of any <see cref="ISchema"/>.
         /// </summary>
@@ -310,38 +300,37 @@ private static Delegate GetMetadataGetterDelegate<TValue>(ISchema schema, int co
             return getter;
         }
 
+        /// <summary>
+        /// Legacy method to get the column index.
+        /// DO NOT USE: use <see cref="GetColumnOrNull"/> instead.
+        /// </summary>
+        [BestFriend]
+        internal bool TryGetColumnIndex(string name, out int col)
+        {
+            col = GetColumnOrNull(name)?.Index ?? -1;
+            return col >= 0;
+        }
+
         #region Legacy schema API to be removed
-        public string GetColumnName(int col) => this[col].Name;
+        /// <summary>
+        /// Number of columns in the schema.
+        /// </summary>
+        int ISchema.ColumnCount => _columns.Length;
 
-        public ColumnType GetColumnType(int col) => this[col].Type;
+        string ISchema.GetColumnName(int col) => this[col].Name;
 
-        public IEnumerable<KeyValuePair<string, ColumnType>> GetMetadataTypes(int col)
-        {
-            var meta = this[col].Metadata;
-            if (meta == null)
-                return Enumerable.Empty<KeyValuePair<string, ColumnType>>();
-            return meta.Schema.GetColumns().Select(c => new KeyValuePair<string, ColumnType>(c.column.Name, c.column.Type));
-        }
+        ColumnType ISchema.GetColumnType(int col) => this[col].Type;
 
-        public ColumnType GetMetadataTypeOrNull(string kind, int col)
-        {
-            var meta = this[col].Metadata;
-            if (meta == null)
-                return null;
-            if (meta.Schema.TryGetColumnIndex(kind, out int metaCol))
-                return meta.Schema[metaCol].Type;
-            return null;
-        }
+        IEnumerable<KeyValuePair<string, ColumnType>> ISchema.GetMetadataTypes(int col)
+            => this[col].Metadata.Schema.Select(c => new KeyValuePair<string, ColumnType>(c.Name, c.Type));
 
-        public void GetMetadata<TValue>(string kind, int col, ref TValue value)
-        {
-            var meta = this[col].Metadata;
-            if (meta == null)
-                throw MetadataUtils.ExceptGetMetadata();
-            meta.GetValue(kind, ref value);
-        }
+        ColumnType ISchema.GetMetadataTypeOrNull(string kind, int col)
+            => this[col].Metadata.Schema.GetColumnOrNull(kind)?.Type;
+
+        void ISchema.GetMetadata<TValue>(string kind, int col, ref TValue value)
+            => this[col].Metadata.GetValue(kind, ref value);
 
-        public bool TryGetColumnIndex(string name, out int col)
+        bool ISchema.TryGetColumnIndex(string name, out int col)
         {
             col = GetColumnOrNull(name)?.Index ?? -1;
             return col >= 0;

src/Microsoft.ML.Core/Data/SchemaDebuggerProxy.cs

@@ -39,10 +39,10 @@ public MetadataDebuggerProxy(Schema.Metadata metadata)
         private static List<KeyValuePair<string, object>> BuildValues(Schema.Metadata metadata)
         {
             var result = new List<KeyValuePair<string, object>>();
-            foreach ((var index, var column) in metadata.Schema.GetColumns())
+            foreach (var column in metadata.Schema)
             {
                 var name = column.Name;
-                var value = Utils.MarshalInvoke(GetValue<int>, column.Type.RawType, metadata, index);
+                var value = Utils.MarshalInvoke(GetValue<int>, column.Type.RawType, metadata, column.Index);
                 result.Add(new KeyValuePair<string, object>(name, value));
             }
             return result;

src/Microsoft.ML.Data/Commands/CrossValidationCommand.cs

@@ -298,7 +298,7 @@ private string GetSplitColumn(IChannel ch, IDataView input, ref IDataView output
                 if (group != null && schema.TryGetColumnIndex(group, out index))
                 {
                     // Check if group column key type with known cardinality.
-                    var type = schema.GetColumnType(index);
+                    var type = schema[index].Type;
                     if (type.KeyCount > 0)
                         stratificationColumn = group;
                 }
@@ -322,7 +322,7 @@ private string GetSplitColumn(IChannel ch, IDataView input, ref IDataView output
                 int col;
                 if (!input.Schema.TryGetColumnIndex(stratificationColumn, out col))
                     throw ch.ExceptUserArg(nameof(Arguments.StratificationColumn), "Column '{0}' does not exist", stratificationColumn);
-                var type = input.Schema.GetColumnType(col);
+                var type = input.Schema[col].Type;
                 if (!RangeFilter.IsValidRangeFilterColumnType(ch, type))
                 {
                     ch.Info("Hashing the stratification column");

src/Microsoft.ML.Data/Commands/DataCommand.cs

@@ -165,8 +165,8 @@ protected void SendTelemetryMetric(Dictionary<string, IDataView>[] metricValues)
                         {
                             for (int currentIndex = 0; currentIndex < cursor.Schema.Count; currentIndex++)
                             {
-                                var nameOfMetric = "TLC_" + cursor.Schema.GetColumnName(currentIndex);
-                                var type = cursor.Schema.GetColumnType(currentIndex);
+                                var nameOfMetric = "TLC_" + cursor.Schema[currentIndex].Name;
+                                var type = cursor.Schema[currentIndex].Type;
                                 if (type.IsNumber)
                                 {
                                     var getter = RowCursorUtils.GetGetterAs<double>(NumberType.R8, cursor, currentIndex);

src/Microsoft.ML.Data/Commands/SaveDataCommand.cs

@@ -146,11 +146,11 @@ private void RunCore(IChannel ch)
             {
                 if (!Args.KeepHidden && data.Schema[i].IsHidden)
                     continue;
-                var type = data.Schema.GetColumnType(i);
+                var type = data.Schema[i].Type;
                 if (saver.IsColumnSavable(type))
                     cols.Add(i);
                 else
-                    ch.Info(MessageSensitivity.Schema, "The column '{0}' will not be written as it has unsavable column type.", data.Schema.GetColumnName(i));
+                    ch.Info(MessageSensitivity.Schema, "The column '{0}' will not be written as it has unsavable column type.", data.Schema[i].Name);
             }
             Host.NotSensitive().Check(cols.Count > 0, "No valid columns to save");
 
@@ -203,15 +203,15 @@ public static void SaveDataView(IChannel ch, IDataSaver saver, IDataView view, S
             ch.CheckValue(stream, nameof(stream));
 
             var cols = new List<int>();
-            for (int i = 0; i < view.Schema.ColumnCount; i++)
+            for (int i = 0; i < view.Schema.Count; i++)
             {
                 if (!keepHidden && view.Schema[i].IsHidden)
                     continue;
-                var type = view.Schema.GetColumnType(i);
+                var type = view.Schema[i].Type;
                 if (saver.IsColumnSavable(type))
                     cols.Add(i);
                 else
-                    ch.Info(MessageSensitivity.Schema, "The column '{0}' will not be written as it has unsavable column type.", view.Schema.GetColumnName(i));
+                    ch.Info(MessageSensitivity.Schema, "The column '{0}' will not be written as it has unsavable column type.", view.Schema[i].Name);
             }
 
             ch.Check(cols.Count > 0, "No valid columns to save");

src/Microsoft.ML.Data/Commands/ScoreCommand.cs

@@ -185,13 +185,13 @@ private void RunCore(IChannel ch)
                     continue;
                 if (!(outputAllColumns || ShouldAddColumn(loader.Schema, i, maxScoreId, outputNamesAndLabels)))
                     continue;
-                var type = loader.Schema.GetColumnType(i);
+                var type = loader.Schema[i].Type;
                 if (writer.IsColumnSavable(type))
                     cols.Add(i);
                 else
                 {
                     ch.Warning("The column '{0}' will not be written as it has unsavable column type.",
-                        loader.Schema.GetColumnName(i));
+                        loader.Schema[i].Name);
                 }
             }
 
@@ -217,7 +217,7 @@ private bool ShouldAddColumn(Schema schema, int i, uint scoreSet, bool outputNam
             }
             if (outputNamesAndLabels)
             {
-                switch (schema.GetColumnName(i))
+                switch (schema[i].Name)
                 {
                     case "Label":
                     case "Name":
@@ -227,7 +227,7 @@ private bool ShouldAddColumn(Schema schema, int i, uint scoreSet, bool outputNam
                         break;
                 }
             }
-            if (Args.OutputColumn != null && Array.FindIndex(Args.OutputColumn, schema.GetColumnName(i).Equals) >= 0)
+            if (Args.OutputColumn != null && Array.FindIndex(Args.OutputColumn, schema[i].Name.Equals) >= 0)
                 return true;
             return false;
         }