Github user davies commented on a diff in the pull request:
https://github.com/apache/spark/pull/6354#discussion_r30934342
--- Diff: python/pyspark/mllib/linalg.py ---
@@ -163,6 +163,59 @@ def simpleString(self):
return "vector"
+class MatrixUDT(UserDefinedType):
+ """
+ SQL user-defined type (UDT) for Matrix.
+ """
+
+ @classmethod
+ def sqlType(cls):
+ return StructType([
+ StructField("type", ByteType(), False),
+ StructField("numRows", IntegerType(), False),
+ StructField("numCols", IntegerType(), False),
+ StructField("colPtrs", ArrayType(IntegerType(), False), True),
+ StructField("rowIndices", ArrayType(IntegerType(), False),
True),
+ StructField("values", ArrayType(DoubleType(), False), True),
+ StructField("isTransposed", BooleanType(), False)])
+
+ @classmethod
+ def module(cls):
+ return "pyspark.mllib.linalg"
+
+ @classmethod
+ def scalaUDT(cls):
+ return "org.apache.spark.mllib.linalg.MatrixUDT"
+
+ def serialize(self, obj):
+ if isinstance(obj, SparseMatrix):
+ colPtrs = [int(i) for i in obj.colPtrs]
+ rowIndices = [int(i) for i in obj.rowIndices]
+ values = [float(v) for v in obj.values]
+ return (0, obj.numRows, obj.numCols, colPtrs,
+ rowIndices, values, bool(obj.isTransposed))
+ elif isinstance(obj, DenseMatrix):
+ values = [float(v) for v in obj.values]
+ return (1, obj.numRows, obj.numCols, None, None, values,
+ bool(obj.isTransposed))
+ else:
+ raise TypeError("cannot serialize %r of type %r" % (obj,
type(obj)))
--- End diff --
the `repr(obj)` could be very long, I think just having `type(obj)` is fine.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
