org.apache.sysml.lops.GroupedAggregate Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of systemml Show documentation
Show all versions of systemml Show documentation
Declarative Machine Learning
/*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/
package org.apache.sysml.lops;
import java.util.HashMap;
import java.util.Map.Entry;
import org.apache.sysml.lops.LopProperties.ExecLocation;
import org.apache.sysml.lops.LopProperties.ExecType;
import org.apache.sysml.lops.compile.JobType;
import org.apache.sysml.parser.Statement;
import org.apache.sysml.parser.Expression.*;
/**
* Lop to perform grouped aggregates
*
*/
public class GroupedAggregate extends Lop
{
private HashMap _inputParams;
private static final String opcode = "groupedagg";
public static final String COMBINEDINPUT = "combinedinput";
private boolean _weights = false;
//spark-specific parameters
private boolean _broadcastGroups = false;
//cp-specific parameters
private int _numThreads = 1;
/**
* Constructor to perform grouped aggregate.
* inputParameterLops <- parameters required to compute different aggregates (hashmap)
* "combinedinput" -- actual data
* "function" -- aggregate function
*
* @param inputParameterLops map of input parameter lops
* @param weights weights
* @param dt data type
* @param vt value type
*/
public GroupedAggregate(
HashMap inputParameterLops, boolean weights,
DataType dt, ValueType vt) {
this(inputParameterLops, dt, vt, ExecType.MR);
_weights = weights;
}
public GroupedAggregate(
HashMap inputParameterLops,
DataType dt, ValueType vt, ExecType et) {
super(Lop.Type.GroupedAgg, dt, vt);
init(inputParameterLops, dt, vt, et);
}
public GroupedAggregate(
HashMap inputParameterLops,
DataType dt, ValueType vt, ExecType et, boolean broadcastGroups) {
super(Lop.Type.GroupedAgg, dt, vt);
init(inputParameterLops, dt, vt, et);
_broadcastGroups = broadcastGroups;
}
public GroupedAggregate(
HashMap inputParameterLops,
DataType dt, ValueType vt, ExecType et, int k) {
super(Lop.Type.GroupedAgg, dt, vt);
init(inputParameterLops, dt, vt, et);
_numThreads = k;
}
private void init(HashMap inputParameterLops,
DataType dt, ValueType vt, ExecType et) {
if ( et == ExecType.MR ) {
/*
* Inputs to ParameterizedBuiltinOp can be in an arbitrary order. However,
* piggybacking (Dag.java:getAggAndOtherInstructions()) expects the first
* input to be the data (named as "combinedinput") on which the aggregate
* needs to be computed. Make sure that "combinedinput" is the first input
* to GroupedAggregate lop.
*/
this.addInput(inputParameterLops.get(COMBINEDINPUT));
inputParameterLops.get(COMBINEDINPUT).addOutput(this);
// process remaining parameters
for ( Entry e : inputParameterLops.entrySet() ) {
String k = e.getKey();
Lop lop = e.getValue();
if ( !k.equalsIgnoreCase(COMBINEDINPUT) ) {
this.addInput(lop);
lop.addOutput(this);
}
}
_inputParams = inputParameterLops;
boolean breaksAlignment = false;
boolean aligner = false;
boolean definesMRJob = true;
lps.addCompatibility(JobType.GROUPED_AGG);
this.lps.setProperties(inputs, et, ExecLocation.MapAndReduce, breaksAlignment, aligner, definesMRJob);
}
else {
boolean breaksAlignment = false;
boolean aligner = false;
boolean definesMRJob = false;
// First, add inputs corresponding to "target" and "groups"
this.addInput(inputParameterLops.get(Statement.GAGG_TARGET));
inputParameterLops.get(Statement.GAGG_TARGET).addOutput(this);
this.addInput(inputParameterLops.get(Statement.GAGG_GROUPS));
inputParameterLops.get(Statement.GAGG_GROUPS).addOutput(this);
// process remaining parameters
for ( Entry e : inputParameterLops.entrySet() ) {
String k = e.getKey();
Lop lop = e.getValue();
if ( !k.equalsIgnoreCase(Statement.GAGG_TARGET) && !k.equalsIgnoreCase(Statement.GAGG_GROUPS) ) {
this.addInput(lop);
lop.addOutput(this);
}
}
_inputParams = inputParameterLops;
lps.addCompatibility(JobType.INVALID);
this.lps.setProperties(inputs, et, ExecLocation.ControlProgram, breaksAlignment, aligner, definesMRJob);
}
}
@Override
public String toString() {
return "Operation = GroupedAggregate";
}
/**
* Function to generate CP Grouped Aggregate Instructions.
*
*/
@Override
public String getInstructions(String output)
throws LopsException
{
StringBuilder sb = new StringBuilder();
sb.append( getExecType() );
sb.append( Lop.OPERAND_DELIMITOR );
sb.append( opcode );
sb.append( Lop.OPERAND_DELIMITOR );
if ( _inputParams.get(Statement.GAGG_TARGET) == null || _inputParams.get(Statement.GAGG_GROUPS) == null || _inputParams.get("fn") == null )
throw new LopsException(this.printErrorLocation() + "Invalid parameters to groupedAggregate -- \"target\", \"groups\", \"fn\" must be provided");
String targetVar = _inputParams.get(Statement.GAGG_TARGET).getOutputParameters().getLabel();
String groupsVar = _inputParams.get(Statement.GAGG_GROUPS).getOutputParameters().getLabel();
sb.append( Statement.GAGG_TARGET );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( targetVar );
sb.append( Lop.OPERAND_DELIMITOR );
sb.append( Statement.GAGG_GROUPS );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( groupsVar );
if ( _inputParams.get(Statement.GAGG_WEIGHTS) != null )
{
sb.append( Lop.OPERAND_DELIMITOR );
sb.append( Statement.GAGG_WEIGHTS );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( _inputParams.get(Statement.GAGG_WEIGHTS).getOutputParameters().getLabel() );
}
// Process all other name=value parameters, which are scalars
String name, valueString;
Lop value;
for(Entry e : _inputParams.entrySet()) {
name = e.getKey();
if ( !name.equalsIgnoreCase(Statement.GAGG_TARGET) && !name.equalsIgnoreCase(Statement.GAGG_GROUPS) && !name.equalsIgnoreCase(Statement.GAGG_WEIGHTS) ) {
value = e.getValue();
valueString = value.prepScalarLabel();
sb.append( OPERAND_DELIMITOR );
sb.append( name );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( valueString );
}
}
if( getExecType()==ExecType.CP ) {
sb.append( OPERAND_DELIMITOR );
sb.append( "k" );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( _numThreads );
}
else if( getExecType()==ExecType.SPARK ) {
sb.append( OPERAND_DELIMITOR );
sb.append( "broadcast" );
sb.append( Lop.NAME_VALUE_SEPARATOR );
sb.append( _broadcastGroups );
}
sb.append( OPERAND_DELIMITOR );
sb.append( prepOutputOperand(output));
return sb.toString();
}
@Override
public String getInstructions(int input_index, int output_index)
{
StringBuilder sb = new StringBuilder();
sb.append( getExecType() );
sb.append( Lop.OPERAND_DELIMITOR );
sb.append( opcode );
sb.append( OPERAND_DELIMITOR );
sb.append( getInputs().get(0).prepInputOperand(input_index));
// get the aggregate function
sb.append( OPERAND_DELIMITOR );
Lop funcLop = _inputParams.get(Statement.GAGG_FN);
sb.append( funcLop.prepScalarInputOperand(getExecType()));
// get the "optional" parameters
if ( _inputParams.get(Statement.GAGG_FN_CM_ORDER) != null ) {
sb.append( OPERAND_DELIMITOR );
Lop orderLop = _inputParams.get(Statement.GAGG_FN_CM_ORDER);
sb.append( orderLop.prepScalarInputOperand(getExecType()));
}
// add output_index to instruction
sb.append( OPERAND_DELIMITOR );
sb.append( prepOutputOperand(output_index) );
sb.append( OPERAND_DELIMITOR );
sb.append( _weights );
sb.append( OPERAND_DELIMITOR );
Lop ngroups = _inputParams.get(Statement.GAGG_NUM_GROUPS);
sb.append( (ngroups!=null)? ngroups.prepScalarInputOperand(getExecType()) : "-1" );
return sb.toString();
}
}