QGIS API Documentation  3.26.3-Buenos Aires (65e4edfdad)
qgsalgorithmrandomextract.cpp
Go to the documentation of this file.
1 /***************************************************************************
2  qgsalgorithmrandomextract.cpp
3  ---------------------
4  begin : December 2019
5  copyright : (C) 2019 by Alexander Bruy
6  email : alexander dot bruy at gmail dot com
7  ***************************************************************************/
8 
9 /***************************************************************************
10  * *
11  * This program is free software; you can redistribute it and/or modify *
12  * it under the terms of the GNU General Public License as published by *
13  * the Free Software Foundation; either version 2 of the License, or *
14  * (at your option) any later version. *
15  * *
16  ***************************************************************************/
17 
19 #include <random>
20 #include <functional>
21 
23 
24 QString QgsRandomExtractAlgorithm::name() const
25 {
26  return QStringLiteral( "randomextract" );
27 }
28 
29 QString QgsRandomExtractAlgorithm::displayName() const
30 {
31  return QObject::tr( "Random extract" );
32 }
33 
34 QStringList QgsRandomExtractAlgorithm::tags() const
35 {
36  return QObject::tr( "extract,filter,random,number,percentage" ).split( ',' );
37 }
38 
39 QString QgsRandomExtractAlgorithm::group() const
40 {
41  return QObject::tr( "Vector selection" );
42 }
43 
44 QString QgsRandomExtractAlgorithm::groupId() const
45 {
46  return QStringLiteral( "vectorselection" );
47 }
48 
49 QString QgsRandomExtractAlgorithm::shortHelpString() const
50 {
51  return QObject::tr( "This algorithm takes a vector layer and generates a new one that contains only a subset "
52  "of the features in the input layer.\n\n"
53  "The subset is defined randomly, using a percentage or count value to define the total number "
54  "of features in the subset." );
55 }
56 
57 QgsRandomExtractAlgorithm *QgsRandomExtractAlgorithm::createInstance() const
58 {
59  return new QgsRandomExtractAlgorithm();
60 }
61 
62 void QgsRandomExtractAlgorithm::initAlgorithm( const QVariantMap & )
63 {
64  addParameter( new QgsProcessingParameterFeatureSource( QStringLiteral( "INPUT" ), QObject::tr( "Input layer" ),
65  QList< int >() << QgsProcessing::TypeVector ) );
66  addParameter( new QgsProcessingParameterEnum( QStringLiteral( "METHOD" ), QObject::tr( "Method" ), QStringList() << QObject::tr( "Number of features" ) << QObject::tr( "Percentage of features" ), false, 0 ) );
67  addParameter( new QgsProcessingParameterNumber( QStringLiteral( "NUMBER" ), QObject::tr( "Number/percentage of features" ),
68  QgsProcessingParameterNumber::Integer, 10, false, 0 ) );
69 
70  addParameter( new QgsProcessingParameterFeatureSink( QStringLiteral( "OUTPUT" ), QObject::tr( "Extracted (random)" ) ) );
71 }
72 
73 QVariantMap QgsRandomExtractAlgorithm::processAlgorithm( const QVariantMap &parameters, QgsProcessingContext &context, QgsProcessingFeedback *feedback )
74 {
75  std::unique_ptr< QgsProcessingFeatureSource > source( parameterAsSource( parameters, QStringLiteral( "INPUT" ), context ) );
76  if ( !source )
77  throw QgsProcessingException( invalidSourceError( parameters, QStringLiteral( "INPUT" ) ) );
78 
79  QString dest;
80  std::unique_ptr< QgsFeatureSink > sink( parameterAsSink( parameters, QStringLiteral( "OUTPUT" ), context, dest, source->fields(),
81  source->wkbType(), source->sourceCrs(), QgsFeatureSink::RegeneratePrimaryKey ) );
82  if ( !sink )
83  throw QgsProcessingException( invalidSinkError( parameters, QStringLiteral( "OUTPUT" ) ) );
84 
85  const int method = parameterAsEnum( parameters, QStringLiteral( "METHOD" ), context );
86  int number = parameterAsInt( parameters, QStringLiteral( "NUMBER" ), context );
87 
88  const long count = source->featureCount();
89 
90  if ( method == 0 )
91  {
92  // number of features
93  if ( number > count )
94  throw QgsProcessingException( QObject::tr( "Selected number is greater than feature count. Choose a lower value and try again." ) );
95  }
96  else
97  {
98  // percentage of features
99  if ( number > 100 )
100  throw QgsProcessingException( QObject::tr( "Percentage can't be greater than 100. Choose a lower value and try again." ) );
101 
102  number = static_cast< int >( std::ceil( number * count / 100 ) );
103  }
104 
105  // initialize random engine
106  std::random_device randomDevice;
107  const std::mt19937 mersenneTwister( randomDevice() );
108  const std::uniform_int_distribution<int> fidsDistribution( 0, count );
109 
110  QVector< QgsFeatureId > fids( number );
111  std::generate( fids.begin(), fids.end(), bind( fidsDistribution, mersenneTwister ) );
112 
113  QHash< QgsFeatureId, int > idsCount;
114  for ( const QgsFeatureId id : fids )
115  {
116  if ( feedback->isCanceled() )
117  {
118  break;
119  }
120 
121  idsCount[ id ] += 1;
122  }
123 
124  const QgsFeatureIds ids = qgis::listToSet( idsCount.keys() );
125  QgsFeatureIterator fit = source->getFeatures( QgsFeatureRequest().setFilterFids( ids ), QgsProcessingFeatureSource::FlagSkipGeometryValidityChecks );
126 
127  QgsFeature f;
128  while ( fit.nextFeature( f ) )
129  {
130  if ( feedback->isCanceled() )
131  {
132  break;
133  }
134 
135  const int count = idsCount.value( f.id() );
136  for ( int i = 0; i < count; ++i )
137  {
138  if ( !sink->addFeature( f, QgsFeatureSink::FastInsert ) )
139  throw QgsProcessingException( writeFeatureError( sink.get(), parameters, QStringLiteral( "OUTPUT" ) ) );
140  }
141  }
142 
143  QVariantMap outputs;
144  outputs.insert( QStringLiteral( "OUTPUT" ), dest );
145  return outputs;
146 }
147 
QgsProcessingParameterNumber
A numeric parameter for processing algorithms.
Definition: qgsprocessingparameters.h:2179
QgsProcessingFeedback
Base class for providing feedback from a processing algorithm.
Definition: qgsprocessingfeedback.h:37
QgsFeedback::isCanceled
bool isCanceled() const SIP_HOLDGIL
Tells whether the operation has been canceled already.
Definition: qgsfeedback.h:67
QgsProcessingFeatureSource::FlagSkipGeometryValidityChecks
@ FlagSkipGeometryValidityChecks
Invalid geometry checks should always be skipped. This flag can be useful for algorithms which always...
Definition: qgsprocessingutils.h:584
QgsProcessingParameterFeatureSource
An input feature source (such as vector layers) parameter for processing algorithms.
Definition: qgsprocessingparameters.h:3057
QgsProcessingParameterFeatureSink
A feature sink output for processing algorithms.
Definition: qgsprocessingparameters.h:3219
QgsFeature::id
QgsFeatureId id
Definition: qgsfeature.h:68
QgsFeatureRequest
This class wraps a request for features to a vector layer (or directly its vector data provider).
Definition: qgsfeaturerequest.h:83
QgsProcessing::TypeVector
@ TypeVector
Tables (i.e. vector layers with or without geometry). When used for a sink this indicates the sink ha...
Definition: qgsprocessing.h:54
QgsProcessingContext
Contains information about the context in which a processing algorithm is executed.
Definition: qgsprocessingcontext.h:46
qgsalgorithmrandomextract.h
QgsFeatureIds
QSet< QgsFeatureId > QgsFeatureIds
Definition: qgsfeatureid.h:37
QgsFeatureSink::RegeneratePrimaryKey
@ RegeneratePrimaryKey
This flag indicates, that a primary key field cannot be guaranteed to be unique and the sink should i...
Definition: qgsfeaturesink.h:55
QgsProcessingParameterNumber::Integer
@ Integer
Integer values.
Definition: qgsprocessingparameters.h:2186
QgsFeatureIterator::nextFeature
bool nextFeature(QgsFeature &f)
Definition: qgsfeatureiterator.h:399
QgsFeature
The feature class encapsulates a single feature including its unique ID, geometry and a list of field...
Definition: qgsfeature.h:55
QgsProcessingParameterEnum
An enum based parameter for processing algorithms, allowing for selection from predefined values.
Definition: qgsprocessingparameters.h:2540
QgsFeatureIterator
Wrapper for iterator of features from vector data provider or vector layer.
Definition: qgsfeatureiterator.h:289
QgsProcessingException
Custom exception class for processing related exceptions.
Definition: qgsexception.h:82
QgsFeatureSink::FastInsert
@ FastInsert
Use faster inserts, at the cost of updating the passed features to reflect changes made at the provid...
Definition: qgsfeaturesink.h:70
QgsFeatureId
qint64 QgsFeatureId
64 bit feature ids negative numbers are used for uncommitted/newly added features
Definition: qgsfeatureid.h:28