QGIS API Documentation 3.40.0-Bratislava (b56115d8743)
Loading...
Searching...
No Matches
qgsalgorithmjoinbynearest.cpp
Go to the documentation of this file.
1/***************************************************************************
2 qgsalgorithmjoinbynearest.cpp
3 ---------------------
4 begin : April 2017
5 copyright : (C) 2017 by Nyall Dawson
6 email : nyall dot dawson 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
20#include "qgslinestring.h"
21#include "qgsspatialindex.h"
22#include <algorithm>
23
25
26QString QgsJoinByNearestAlgorithm::name() const
27{
28 return QStringLiteral( "joinbynearest" );
29}
30
31QString QgsJoinByNearestAlgorithm::displayName() const
32{
33 return QObject::tr( "Join attributes by nearest" );
34}
35
36QStringList QgsJoinByNearestAlgorithm::tags() const
37{
38 return QObject::tr( "join,connect,attributes,values,fields,tables,proximity,closest,neighbour,neighbor,n-nearest,distance" ).split( ',' );
39}
40
41QString QgsJoinByNearestAlgorithm::group() const
42{
43 return QObject::tr( "Vector general" );
44}
45
46QString QgsJoinByNearestAlgorithm::groupId() const
47{
48 return QStringLiteral( "vectorgeneral" );
49}
50
51void QgsJoinByNearestAlgorithm::initAlgorithm( const QVariantMap & )
52{
53 addParameter( new QgsProcessingParameterFeatureSource( QStringLiteral( "INPUT" ),
54 QObject::tr( "Input layer" ) ) );
55 addParameter( new QgsProcessingParameterFeatureSource( QStringLiteral( "INPUT_2" ),
56 QObject::tr( "Input layer 2" ) ) );
57
58 addParameter( new QgsProcessingParameterField( QStringLiteral( "FIELDS_TO_COPY" ),
59 QObject::tr( "Layer 2 fields to copy (leave empty to copy all fields)" ),
60 QVariant(), QStringLiteral( "INPUT_2" ), Qgis::ProcessingFieldParameterDataType::Any,
61 true, true ) );
62
63 addParameter( new QgsProcessingParameterBoolean( QStringLiteral( "DISCARD_NONMATCHING" ),
64 QObject::tr( "Discard records which could not be joined" ),
65 false ) );
66
67 addParameter( new QgsProcessingParameterString( QStringLiteral( "PREFIX" ),
68 QObject::tr( "Joined field prefix" ), QVariant(), false, true ) );
69
70 addParameter( new QgsProcessingParameterNumber( QStringLiteral( "NEIGHBORS" ),
71 QObject::tr( "Maximum nearest neighbors" ), Qgis::ProcessingNumberParameterType::Integer, 1, false, 1 ) );
72
73 addParameter( new QgsProcessingParameterDistance( QStringLiteral( "MAX_DISTANCE" ),
74 QObject::tr( "Maximum distance" ), QVariant(), QStringLiteral( "INPUT" ), true, 0 ) );
75
76 addParameter( new QgsProcessingParameterFeatureSink( QStringLiteral( "OUTPUT" ), QObject::tr( "Joined layer" ), Qgis::ProcessingSourceType::VectorAnyGeometry, QVariant(), true, true ) );
77
78 std::unique_ptr< QgsProcessingParameterFeatureSink > nonMatchingSink = std::make_unique< QgsProcessingParameterFeatureSink >(
79 QStringLiteral( "NON_MATCHING" ), QObject::tr( "Unjoinable features from first layer" ), Qgis::ProcessingSourceType::VectorAnyGeometry, QVariant(), true, false );
80 // TODO GUI doesn't support advanced outputs yet
81 //nonMatchingSink->setFlags(nonMatchingSink->flags() | Qgis::ProcessingParameterFlag::Advanced );
82 addParameter( nonMatchingSink.release() );
83
84 addOutput( new QgsProcessingOutputNumber( QStringLiteral( "JOINED_COUNT" ), QObject::tr( "Number of joined features from input table" ) ) );
85 addOutput( new QgsProcessingOutputNumber( QStringLiteral( "UNJOINABLE_COUNT" ), QObject::tr( "Number of unjoinable features from input table" ) ) );
86}
87
88QString QgsJoinByNearestAlgorithm::shortHelpString() const
89{
90 return QObject::tr( "This algorithm takes an input vector layer and creates a new vector layer that is an extended version of the "
91 "input one, with additional attributes in its attribute table.\n\n"
92 "The additional attributes and their values are taken from a second vector layer, where features are joined "
93 "by finding the closest features from each layer. By default only the single nearest feature is joined,"
94 "but optionally the join can use the n-nearest neighboring features instead. If multiple features are found "
95 "with identical distances these will all be returned (even if the total number of features exceeds the specified "
96 "maximum feature count).\n\n"
97 "If a maximum distance is specified, then only features which are closer than this distance "
98 "will be matched.\n\n"
99 "The output features will contain the selected attributes from the nearest feature, "
100 "along with new attributes for the distance to the near feature, the index of the feature, "
101 "and the coordinates of the closest point on the input feature (feature_x, feature_y) "
102 "to the matched nearest feature, and the coordinates of the closet point on the matched feature "
103 "(nearest_x, nearest_y).\n\n"
104 "This algorithm uses purely Cartesian calculations for distance, and does not consider "
105 "geodetic or ellipsoid properties when determining feature proximity." );
106}
107
108QString QgsJoinByNearestAlgorithm::shortDescription() const
109{
110 return QObject::tr( "Joins a layer to another layer, using the closest features (nearest neighbors)." );
111}
112
113Qgis::ProcessingAlgorithmDocumentationFlags QgsJoinByNearestAlgorithm::documentationFlags() const
114{
116}
117
118QgsJoinByNearestAlgorithm *QgsJoinByNearestAlgorithm::createInstance() const
119{
120 return new QgsJoinByNearestAlgorithm();
121}
122
123QVariantMap QgsJoinByNearestAlgorithm::processAlgorithm( const QVariantMap &parameters, QgsProcessingContext &context, QgsProcessingFeedback *feedback )
124{
125 const int neighbors = parameterAsInt( parameters, QStringLiteral( "NEIGHBORS" ), context );
126 const bool discardNonMatching = parameterAsBoolean( parameters, QStringLiteral( "DISCARD_NONMATCHING" ), context );
127 const double maxDistance = parameters.value( QStringLiteral( "MAX_DISTANCE" ) ).isValid() ? parameterAsDouble( parameters, QStringLiteral( "MAX_DISTANCE" ), context ) : std::numeric_limits< double >::quiet_NaN();
128 std::unique_ptr< QgsProcessingFeatureSource > input( parameterAsSource( parameters, QStringLiteral( "INPUT" ), context ) );
129 if ( !input )
130 throw QgsProcessingException( invalidSourceError( parameters, QStringLiteral( "INPUT" ) ) );
131
132 std::unique_ptr< QgsProcessingFeatureSource > input2( parameterAsSource( parameters, QStringLiteral( "INPUT_2" ), context ) );
133 if ( !input2 )
134 throw QgsProcessingException( invalidSourceError( parameters, QStringLiteral( "INPUT_2" ) ) );
135
136 const bool sameSourceAndTarget = parameters.value( QStringLiteral( "INPUT" ) ) == parameters.value( QStringLiteral( "INPUT_2" ) );
137
138 const QString prefix = parameterAsString( parameters, QStringLiteral( "PREFIX" ), context );
139 const QStringList fieldsToCopy = parameterAsStrings( parameters, QStringLiteral( "FIELDS_TO_COPY" ), context );
140
141 QgsFields outFields2;
142 QgsAttributeList fields2Indices;
143 if ( fieldsToCopy.empty() )
144 {
145 outFields2 = input2->fields();
146 fields2Indices.reserve( outFields2.count() );
147 for ( int i = 0; i < outFields2.count(); ++i )
148 {
149 fields2Indices << i;
150 }
151 }
152 else
153 {
154 fields2Indices.reserve( fieldsToCopy.count() );
155 for ( const QString &field : fieldsToCopy )
156 {
157 const int index = input2->fields().lookupField( field );
158 if ( index >= 0 )
159 {
160 fields2Indices << index;
161 outFields2.append( input2->fields().at( index ) );
162 }
163 }
164 }
165
166 if ( !prefix.isEmpty() )
167 {
168 for ( int i = 0; i < outFields2.count(); ++i )
169 {
170 outFields2.rename( i, prefix + outFields2[ i ].name() );
171 }
172 }
173
174 const QgsAttributeList fields2Fetch = fields2Indices;
175
176 QgsFields outFields = QgsProcessingUtils::combineFields( input->fields(), outFields2 );
177
178 QgsFields resultFields;
179 resultFields.append( QgsField( QStringLiteral( "n" ), QMetaType::Type::Int ) );
180 resultFields.append( QgsField( QStringLiteral( "distance" ), QMetaType::Type::Double ) );
181 resultFields.append( QgsField( QStringLiteral( "feature_x" ), QMetaType::Type::Double ) );
182 resultFields.append( QgsField( QStringLiteral( "feature_y" ), QMetaType::Type::Double ) );
183 resultFields.append( QgsField( QStringLiteral( "nearest_x" ), QMetaType::Type::Double ) );
184 resultFields.append( QgsField( QStringLiteral( "nearest_y" ), QMetaType::Type::Double ) );
185 outFields = QgsProcessingUtils::combineFields( outFields, resultFields );
186
187 QString dest;
188 std::unique_ptr< QgsFeatureSink > sink( parameterAsSink( parameters, QStringLiteral( "OUTPUT" ), context, dest, outFields,
189 input->wkbType(), input->sourceCrs(), QgsFeatureSink::RegeneratePrimaryKey ) );
190 if ( parameters.value( QStringLiteral( "OUTPUT" ) ).isValid() && !sink )
191 throw QgsProcessingException( invalidSinkError( parameters, QStringLiteral( "OUTPUT" ) ) );
192
193 QString destNonMatching1;
194 std::unique_ptr< QgsFeatureSink > sinkNonMatching1( parameterAsSink( parameters, QStringLiteral( "NON_MATCHING" ), context, destNonMatching1, input->fields(),
195 input->wkbType(), input->sourceCrs(), QgsFeatureSink::RegeneratePrimaryKey ) );
196 if ( parameters.value( QStringLiteral( "NON_MATCHING" ) ).isValid() && !sinkNonMatching1 )
197 throw QgsProcessingException( invalidSinkError( parameters, QStringLiteral( "NON_MATCHING" ) ) );
198
199 // make spatial index
200 const QgsFeatureIterator f2 = input2->getFeatures( QgsFeatureRequest().setDestinationCrs( input->sourceCrs(), context.transformContext() ).setSubsetOfAttributes( fields2Fetch ) );
201 QHash< QgsFeatureId, QgsAttributes > input2AttributeCache;
202 double step = input2->featureCount() > 0 ? 50.0 / input2->featureCount() : 1;
203 int i = 0;
204 const QgsSpatialIndex index( f2, [&]( const QgsFeature & f )->bool
205 {
206 i++;
207 if ( feedback->isCanceled() )
208 return false;
209
210 feedback->setProgress( i * step );
211
212 if ( !f.hasGeometry() )
213 return true;
214
215 // only keep selected attributes
216 QgsAttributes attributes;
217 const int attributeCount = f.attributeCount();
218 for ( int j = 0; j < attributeCount; ++j )
219 {
220 if ( ! fields2Indices.contains( j ) )
221 continue;
222 attributes << f.attribute( j );
223 }
224 input2AttributeCache.insert( f.id(), attributes );
225
226 return true;
228
229 QgsFeature f;
230
231 // create extra null attributes for non-matched records (the +2 is for the "n" and "distance", and start/end x/y fields)
232 QgsAttributes nullMatch;
233 nullMatch.reserve( fields2Indices.size() + 6 );
234 for ( int i = 0; i < fields2Indices.count() + 6; ++i )
235 nullMatch << QVariant();
236
237 long long joinedCount = 0;
238 long long unjoinedCount = 0;
239
240 // Create output vector layer with additional attributes
241 step = input->featureCount() > 0 ? 50.0 / input->featureCount() : 1;
242 QgsFeatureIterator features = input->getFeatures();
243 i = 0;
244 while ( features.nextFeature( f ) )
245 {
246 i++;
247 if ( feedback->isCanceled() )
248 {
249 break;
250 }
251
252 feedback->setProgress( 50 + i * step );
253
254 if ( !f.hasGeometry() )
255 {
256 unjoinedCount++;
257 if ( sinkNonMatching1 )
258 {
259 if ( !sinkNonMatching1->addFeature( f, QgsFeatureSink::FastInsert ) )
260 throw QgsProcessingException( writeFeatureError( sinkNonMatching1.get(), parameters, QStringLiteral( "NON_MATCHING" ) ) );
261 }
262 if ( sink && !discardNonMatching )
263 {
264 QgsAttributes attr = f.attributes();
265 attr.append( nullMatch );
266 f.setAttributes( attr );
267 if ( !sink->addFeature( f, QgsFeatureSink::FastInsert ) )
268 throw QgsProcessingException( writeFeatureError( sink.get(), parameters, QStringLiteral( "OUTPUT" ) ) );
269 }
270 }
271 else
272 {
273 // note - if using same source as target, we have to get one extra neighbor, since the first match will be the input feature
274
275 // if the user didn't specify a distance (isnan), then use 0 for nearestNeighbor() parameter
276 // if the user specified 0 exactly, then use the smallest positive double value instead
277 const double searchDistance = std::isnan( maxDistance ) ? 0 : std::max( std::numeric_limits<double>::min(), maxDistance );
278 const QList< QgsFeatureId > nearest = index.nearestNeighbor( f.geometry(), neighbors + ( sameSourceAndTarget ? 1 : 0 ), searchDistance );
279
280 if ( nearest.count() > neighbors + ( sameSourceAndTarget ? 1 : 0 ) )
281 {
282 feedback->pushInfo( QObject::tr( "Multiple matching features found at same distance from search feature, found %n feature(s) instead of %1", nullptr, nearest.count() - ( sameSourceAndTarget ? 1 : 0 ) ).arg( neighbors ) );
283 }
284 QgsFeature out;
285 out.setGeometry( f.geometry() );
286 int j = 0;
287 for ( const QgsFeatureId id : nearest )
288 {
289 if ( sameSourceAndTarget && id == f.id() )
290 continue; // don't match to same feature if using a single input table
291 j++;
292 if ( sink )
293 {
294 QgsAttributes attr = f.attributes();
295 attr.append( input2AttributeCache.value( id ) );
296 attr.append( j );
297
298 const QgsGeometry closestLine = f.geometry().shortestLine( index.geometry( id ) );
299 if ( const QgsLineString *line = qgsgeometry_cast< const QgsLineString *>( closestLine.constGet() ) )
300 {
301 attr.append( line->length() );
302 attr.append( line->startPoint().x() );
303 attr.append( line->startPoint().y() );
304 attr.append( line->endPoint().x() );
305 attr.append( line->endPoint().y() );
306 }
307 else
308 {
309 attr.append( QVariant() ); //distance
310 attr.append( QVariant() ); //start x
311 attr.append( QVariant() ); //start y
312 attr.append( QVariant() ); //end x
313 attr.append( QVariant() ); //end y
314 }
315 out.setAttributes( attr );
316 if ( !sink->addFeature( out, QgsFeatureSink::FastInsert ) )
317 throw QgsProcessingException( writeFeatureError( sink.get(), parameters, QStringLiteral( "OUTPUT" ) ) );
318 }
319 }
320 if ( j > 0 )
321 joinedCount++;
322 else
323 {
324 if ( sinkNonMatching1 )
325 {
326 if ( !sinkNonMatching1->addFeature( f, QgsFeatureSink::FastInsert ) )
327 throw QgsProcessingException( writeFeatureError( sinkNonMatching1.get(), parameters, QStringLiteral( "NON_MATCHING" ) ) );
328 }
329 if ( !discardNonMatching && sink )
330 {
331 QgsAttributes attr = f.attributes();
332 attr.append( nullMatch );
333 f.setAttributes( attr );
334 if ( !sink->addFeature( f, QgsFeatureSink::FastInsert ) )
335 throw QgsProcessingException( writeFeatureError( sink.get(), parameters, QStringLiteral( "OUTPUT" ) ) );
336 }
337 unjoinedCount++;
338 }
339 }
340 }
341
342 QVariantMap outputs;
343 outputs.insert( QStringLiteral( "JOINED_COUNT" ), joinedCount );
344 outputs.insert( QStringLiteral( "UNJOINABLE_COUNT" ), unjoinedCount );
345 if ( sink )
346 outputs.insert( QStringLiteral( "OUTPUT" ), dest );
347 if ( sinkNonMatching1 )
348 outputs.insert( QStringLiteral( "NON_MATCHING" ), destNonMatching1 );
349 return outputs;
350}
351
352
@ VectorAnyGeometry
Any vector layer with geometry.
@ RegeneratesPrimaryKey
Algorithm always drops any existing primary keys or FID values and regenerates them in outputs.
QFlags< ProcessingAlgorithmDocumentationFlag > ProcessingAlgorithmDocumentationFlags
Flags describing algorithm behavior for documentation purposes.
Definition qgis.h:3359
A vector of attributes.
Wrapper for iterator of features from vector data provider or vector layer.
bool nextFeature(QgsFeature &f)
Fetch next feature and stores in f, returns true on success.
This class wraps a request for features to a vector layer (or directly its vector data provider).
QgsFeatureRequest & setSubsetOfAttributes(const QgsAttributeList &attrs)
Set a subset of attributes that will be fetched.
@ FastInsert
Use faster inserts, at the cost of updating the passed features to reflect changes made at the provid...
@ RegeneratePrimaryKey
This flag indicates, that a primary key field cannot be guaranteed to be unique and the sink should i...
The feature class encapsulates a single feature including its unique ID, geometry and a list of field...
Definition qgsfeature.h:58
QgsFeatureId id
Definition qgsfeature.h:66
void setAttributes(const QgsAttributes &attrs)
Sets the feature's attributes.
int attributeCount() const
Returns the number of attributes attached to the feature.
bool hasGeometry() const
Returns true if the feature has an associated geometry.
Q_INVOKABLE QVariant attribute(const QString &name) const
Lookup attribute value by attribute name.
void setGeometry(const QgsGeometry &geometry)
Set the feature's geometry.
bool isCanceled() const
Tells whether the operation has been canceled already.
Definition qgsfeedback.h:53
void setProgress(double progress)
Sets the current progress for the feedback object.
Definition qgsfeedback.h:61
Encapsulate a field in an attribute table or data source.
Definition qgsfield.h:53
Container of fields for a vector layer.
Definition qgsfields.h:46
bool append(const QgsField &field, Qgis::FieldOrigin origin=Qgis::FieldOrigin::Provider, int originIndex=-1)
Appends a field.
Definition qgsfields.cpp:69
int count
Definition qgsfields.h:50
Q_INVOKABLE int lookupField(const QString &fieldName) const
Looks up field's index from the field name.
bool rename(int fieldIdx, const QString &name)
Renames a name of field.
A geometry is the spatial representation of a feature.
const QgsAbstractGeometry * constGet() const
Returns a non-modifiable (const) reference to the underlying abstract geometry primitive.
QgsGeometry shortestLine(const QgsGeometry &other) const
Returns the shortest line joining this geometry to another geometry.
Line string geometry type, with support for z-dimension and m-values.
Contains information about the context in which a processing algorithm is executed.
QgsCoordinateTransformContext transformContext() const
Returns the coordinate transform context.
Custom exception class for processing related exceptions.
Base class for providing feedback from a processing algorithm.
virtual void pushInfo(const QString &info)
Pushes a general informational message from the algorithm.
A numeric output for processing algorithms.
A boolean parameter for processing algorithms.
A double numeric parameter for distance values.
A feature sink output for processing algorithms.
An input feature source (such as vector layers) parameter for processing algorithms.
A vector layer or feature source field parameter for processing algorithms.
A numeric parameter for processing algorithms.
A string parameter for processing algorithms.
static QgsFields combineFields(const QgsFields &fieldsA, const QgsFields &fieldsB, const QString &fieldsBPrefix=QString())
Combines two field lists, avoiding duplicate field names (in a case-insensitive manner).
A spatial index for QgsFeature objects.
@ FlagStoreFeatureGeometries
Indicates that the spatial index should also store feature geometries. This requires more memory,...
qint64 QgsFeatureId
64 bit feature ids negative numbers are used for uncommitted/newly added features
QList< int > QgsAttributeList
Definition qgsfield.h:27