private ReadOnlyMemory <char> GetSlotName(int index, VBuffer <ReadOnlyMemory <char> > slotNames)
{
_env.Assert(0 <= index && index < slotNames.Length);
var slotName = slotNames.GetItemOrDefault(index);
return(slotName.IsEmpty
? new ReadOnlyMemory <char>($"f{index}".ToCharArray())
: slotName);
}
private ReadOnlyMemory <char> GetSlotName(int index, VBuffer <ReadOnlyMemory <char> > slotNames)
{
var count = slotNames.GetValues().Length;
_env.Assert(count > index || count == 0 && slotNames.Length > index);
var slotName = slotNames.GetItemOrDefault(index);
return(slotName.IsEmpty
? new ReadOnlyMemory <char>($"f{index}".ToCharArray())
: slotName);
}
/// <summary>
/// Gets the mapping from T into a StringBuilder representation, using various heuristics.
/// This StringBuilder representation will be a component of the composed KeyValues for the
/// hash outputs.
/// </summary>
public static ValueMapper <T, StringBuilder> GetSimpleMapper <T>(ISchema schema, int col)
{
Contracts.AssertValue(schema);
Contracts.Assert(0 <= col && col < schema.ColumnCount);
var type = schema.GetColumnType(col).ItemType;
Contracts.Assert(type.RawType == typeof(T));
var conv = Conversion.Conversions.Instance;
// First: if not key, then get the standard string converison.
if (!type.IsKey)
{
return(conv.GetStringConversion <T>(type));
}
bool identity;
// Second choice: if key, utilize the KeyValues metadata for that key, if it has one and is text.
if (schema.HasKeyNames(col, type.KeyCount))
{
// REVIEW: Non-textual KeyValues are certainly possible. Should we handle them?
// Get the key names.
VBuffer <ReadOnlyMemory <char> > keyValues = default;
schema.GetMetadata(MetadataUtils.Kinds.KeyValues, col, ref keyValues);
ReadOnlyMemory <char> value = default;
// REVIEW: We could optimize for identity, but it's probably not worthwhile.
var keyMapper = conv.GetStandardConversion <T, uint>(type, NumberType.U4, out identity);
return
((ref T src, ref StringBuilder dst) =>
{
ClearDst(ref dst);
uint intermediate = 0;
keyMapper(ref src, ref intermediate);
if (intermediate == 0)
{
return;
}
keyValues.GetItemOrDefault((int)(intermediate - 1), ref value);
dst.AppendMemory(value);
});
}
// Third choice: just use the key value itself, subject to offsetting by the min.
return(conv.GetKeyStringConversion <T>(type.AsKey));
}
public override Delegate[] CreateGetters(IRow input, Func <int, bool> activeOutput, out Action disposer)
{
disposer = null;
var getters = new Delegate[3];
if (!activeOutput(ClusterIdCol) && !activeOutput(SortedClusterCol) && !activeOutput(SortedClusterScoreCol))
{
return(getters);
}
long cachedPosition = -1;
VBuffer <Single> scores = default(VBuffer <Single>);
var scoresArr = new Single[_numClusters];
int[] sortedIndices = new int[_numClusters];
var scoreGetter = input.GetGetter <VBuffer <Single> >(ScoreIndex);
Action updateCacheIfNeeded =
() =>
{
if (cachedPosition != input.Position)
{
scoreGetter(ref scores);
scores.CopyTo(scoresArr);
int j = 0;
foreach (var index in Enumerable.Range(0, scoresArr.Length).OrderBy(i => scoresArr[i]))
{
sortedIndices[j++] = index;
}
cachedPosition = input.Position;
}
};
if (activeOutput(ClusterIdCol))
{
ValueGetter <uint> assignedFn =
(ref uint dst) =>
{
updateCacheIfNeeded();
dst = (uint)sortedIndices[0] + 1;
};
getters[ClusterIdCol] = assignedFn;
}
if (activeOutput(SortedClusterScoreCol))
{
ValueGetter <VBuffer <Single> > topKScoresFn =
(ref VBuffer <Single> dst) =>
{
updateCacheIfNeeded();
var values = dst.Values;
if (Utils.Size(values) < _numClusters)
{
values = new Single[_numClusters];
}
for (int i = 0; i < _numClusters; i++)
{
values[i] = scores.GetItemOrDefault(sortedIndices[i]);
}
dst = new VBuffer <Single>(_numClusters, values);
};
getters[SortedClusterScoreCol] = topKScoresFn;
}
if (activeOutput(SortedClusterCol))
{
ValueGetter <VBuffer <uint> > topKClassesFn =
(ref VBuffer <uint> dst) =>
{
updateCacheIfNeeded();
var values = dst.Values;
if (Utils.Size(values) < _numClusters)
{
values = new uint[_numClusters];
}
for (int i = 0; i < _numClusters; i++)
{
values[i] = (uint)sortedIndices[i] + 1;
}
dst = new VBuffer <uint>(_numClusters, values);
};
getters[SortedClusterCol] = topKClassesFn;
}
return(getters);
}