/*
- * Jalview - A Sequence Alignment Editor and Viewer (Development Version 2.4.1)
- * Copyright (C) 2009 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
+ * This file is part of Jalview.
*
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
+ * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.analysis;
-import java.util.*;
-
-import jalview.datamodel.*;
-import jalview.util.*;
+import jalview.analysis.scoremodels.PIDModel;
+import jalview.analysis.scoremodels.SimilarityParams;
+import jalview.bin.ApplicationSingletonProvider;
+import jalview.bin.ApplicationSingletonProvider.ApplicationSingletonI;
+import jalview.datamodel.AlignmentAnnotation;
+import jalview.datamodel.AlignmentI;
+import jalview.datamodel.AlignmentOrder;
+import jalview.datamodel.SequenceFeature;
+import jalview.datamodel.SequenceGroup;
+import jalview.datamodel.SequenceI;
+import jalview.datamodel.SequenceNode;
+import jalview.util.QuickSort;
+
+import java.util.ArrayList;
+import java.util.BitSet;
+import java.util.Collections;
+import java.util.Iterator;
+import java.util.List;
/**
* Routines for manipulating the order of a multiple sequence alignment TODO:
* from the first tobesorted position in the alignment. e.g. (a,tb2,b,tb1,c,tb3
* becomes a,tb1,tb2,tb3,b,c)
*/
-public class AlignmentSorter
+public class AlignmentSorter implements ApplicationSingletonI
{
- static boolean sortIdAscending = true;
- static int lastGroupHash = 0;
+ private AlignmentSorter()
+ {
+ // private singleton
+ }
+
+ public static AlignmentSorter getInstance()
+ {
+ return (AlignmentSorter) ApplicationSingletonProvider
+ .getInstance(AlignmentSorter.class);
+ }
+
+ /**
+ * types of feature ordering: Sort by score : average score - or total score -
+ * over all features in region Sort by feature label text: (or if null -
+ * feature type text) - numerical or alphabetical Sort by feature density:
+ * based on counts - ignoring individual text or scores for each feature
+ */
+ public static final String FEATURE_SCORE = "average_score";
+
+ public static final String FEATURE_LABEL = "text";
+
+ public static final String FEATURE_DENSITY = "density";
+
+ /*
+ * todo: refactor searches to follow a basic pattern: (search property, last
+ * search state, current sort direction)
+ */
+ boolean sortIdAscending = true;
+
+ int lastGroupHash = 0;
- static boolean sortGroupAscending = true;
+ boolean sortGroupAscending = true;
- static AlignmentOrder lastOrder = null;
+ AlignmentOrder lastOrder = null;
- static boolean sortOrderAscending = true;
+ boolean sortOrderAscending = true;
- static NJTree lastTree = null;
+ TreeModel lastTree = null;
- static boolean sortTreeAscending = true;
+ boolean sortTreeAscending = true;
/**
- * last Annotation Label used by sortByScore
+ * last Annotation Label used for sort by Annotation score
*/
- private static String lastSortByScore;
+ private String lastSortByAnnotation;
/**
- * compact representation of last arguments to SortByFeatureScore
+ * string hash of last arguments to sortByFeature (sort order toggles if this
+ * is unchanged between sorts)
*/
- private static String lastSortByFeatureScore;
+ private String sortByFeatureCriteria;
- private static boolean sortLengthAscending;
+ private boolean sortByFeatureAscending = true;
+
+ private boolean sortLengthAscending;
+
+ private static boolean sortEValueAscending;
+
+ private static boolean sortBitScoreAscending;
/**
- * Sort by Percentage Identity w.r.t. s
- *
- * @param align
- * AlignmentI
- * @param s
- * SequenceI
- * @param tosort
- * sequences from align that are to be sorted.
- */
- public static void sortByPID(AlignmentI align, SequenceI s,
- SequenceI[] tosort)
- {
- sortByPID(align,s,tosort,0,-1);
- }
- /**
- * Sort by Percentage Identity w.r.t. s
+ * Sorts sequences in the alignment by Percentage Identity with the given
+ * reference sequence, sorting the highest identity to the top
*
* @param align
- * AlignmentI
+ * AlignmentI
* @param s
- * SequenceI
- * @param tosort
- * sequences from align that are to be sorted.
- * @param start start column (0 for beginning
+ * SequenceI
* @param end
*/
- public static void sortByPID(AlignmentI align, SequenceI s,
- SequenceI[] tosort,int start, int end)
+ public static void sortByPID(AlignmentI align, SequenceI s)
{
int nSeq = align.getHeight();
float[] scores = new float[nSeq];
SequenceI[] seqs = new SequenceI[nSeq];
+ String refSeq = s.getSequenceAsString();
+ SimilarityParams pidParams = new SimilarityParams(true, true, true,
+ true);
for (int i = 0; i < nSeq; i++)
{
- scores[i] = Comparison.PID(align.getSequenceAt(i)
- .getSequenceAsString(), s.getSequenceAsString());
+ scores[i] = (float) PIDModel.computePID(
+ align.getSequenceAt(i).getSequenceAsString(), refSeq,
+ pidParams);
seqs[i] = align.getSequenceAt(i);
}
- QuickSort.sort(scores, 0, scores.length - 1, seqs);
-
+ QuickSort.sort(scores, seqs);
setReverseOrder(align, seqs);
}
/**
- * Reverse the order of the sort
+ * Sorts by ID. Numbers are sorted before letters.
*
* @param align
- * DOCUMENT ME!
- * @param seqs
- * DOCUMENT ME!
+ * The alignment object to sort
*/
- private static void setReverseOrder(AlignmentI align, SequenceI[] seqs)
+ public static void sortByID(AlignmentI align)
{
- int nSeq = seqs.length;
-
- int len = 0;
+ int nSeq = align.getHeight();
- if ((nSeq % 2) == 0)
- {
- len = nSeq / 2;
- }
- else
- {
- len = (nSeq + 1) / 2;
- }
+ String[] ids = new String[nSeq];
+ SequenceI[] seqs = new SequenceI[nSeq];
- // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work
- for (int i = 0; i < len; i++)
+ for (int i = 0; i < nSeq; i++)
{
- // SequenceI tmp = seqs[i];
- align.getSequences().setElementAt(seqs[nSeq - i - 1], i);
- align.getSequences().setElementAt(seqs[i], nSeq - i - 1);
+ ids[i] = align.getSequenceAt(i).getName();
+ seqs[i] = align.getSequenceAt(i);
}
- }
- /**
- * Sets the Alignment object with the given sequences
- *
- * @param align
- * Alignment object to be updated
- * @param tmp
- * sequences as a vector
- */
- private static void setOrder(AlignmentI align, Vector tmp)
- {
- setOrder(align, vectorSubsetToArray(tmp, align.getSequences()));
+ QuickSort.sort(ids, seqs);
+ AlignmentSorter as = getInstance();
+ as.sortIdAscending = !as.sortIdAscending;
+ set(align, seqs, as.sortIdAscending);
}
/**
- * Sets the Alignment object with the given sequences
+ * Sorts by sequence length
*
* @param align
- * DOCUMENT ME!
- * @param seqs
- * sequences as an array
+ * The alignment object to sort
*/
- public static void setOrder(AlignmentI align, SequenceI[] seqs)
+ public static void sortByLength(AlignmentI align)
{
- // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work
- Vector algn = align.getSequences();
- Vector tmp = new Vector();
+ int nSeq = align.getHeight();
- for (int i = 0; i < seqs.length; i++)
- {
- if (algn.contains(seqs[i]))
- {
- tmp.addElement(seqs[i]);
- }
- }
+ float[] length = new float[nSeq];
+ SequenceI[] seqs = new SequenceI[nSeq];
- algn.removeAllElements();
- // User may have hidden seqs, then clicked undo or redo
- for (int i = 0; i < tmp.size(); i++)
+ for (int i = 0; i < nSeq; i++)
{
- algn.addElement(tmp.elementAt(i));
+ seqs[i] = align.getSequenceAt(i);
+ length[i] = (seqs[i].getEnd() - seqs[i].getStart());
}
+ QuickSort.sort(length, seqs);
+ AlignmentSorter as = getInstance();
+ as.sortLengthAscending = !as.sortLengthAscending;
+ set(align, seqs, as.sortLengthAscending);
}
/**
- * Sorts by ID. Numbers are sorted before letters.
+ * Sorts by sequence evalue. Currently moves all sequences without an evalue to
+ * the top of the alignment.
*
* @param align
* The alignment object to sort
*/
- public static void sortByID(AlignmentI align)
+ public static void sortByEValue(AlignmentI align)
{
int nSeq = align.getHeight();
- String[] ids = new String[nSeq];
+ double[] evalue = new double[nSeq];
SequenceI[] seqs = new SequenceI[nSeq];
for (int i = 0; i < nSeq; i++)
{
- ids[i] = align.getSequenceAt(i).getName();
seqs[i] = align.getSequenceAt(i);
+ AlignmentAnnotation[] ann = seqs[i].getAnnotation("Search Scores");
+ if (ann != null)
+ {
+ evalue[i] = ann[0].getEValue();
+ }
+ else
+ {
+ evalue[i] = -1;
+ }
}
- QuickSort.sort(ids, seqs);
+ QuickSort.sort(evalue, seqs);
- if (sortIdAscending)
+ if (sortEValueAscending)
{
setReverseOrder(align, seqs);
}
setOrder(align, seqs);
}
- sortIdAscending = !sortIdAscending;
+ sortEValueAscending = !sortEValueAscending;
}
+
/**
- * Sorts by sequence length
+ * Sorts by sequence bit score. Currently moves all sequences without a bit
+ * score to the top of the alignment
*
* @param align
* The alignment object to sort
*/
- public static void sortByLength(AlignmentI align)
+ public static void sortByBitScore(AlignmentI align)
{
int nSeq = align.getHeight();
- float[] length = new float[nSeq];
+ double[] score = new double[nSeq];
SequenceI[] seqs = new SequenceI[nSeq];
-
+
for (int i = 0; i < nSeq; i++)
{
seqs[i] = align.getSequenceAt(i);
- length[i] = (float) (seqs[i].getEnd()-seqs[i].getStart());
+ AlignmentAnnotation[] ann = seqs[i].getAnnotation("Search Scores");
+ if (ann != null)
+ {
+ score[i] = ann[0].getEValue();
+ }
+ else
+ {
+ score[i] = -1;
+ }
}
- QuickSort.sort(length, seqs);
+ QuickSort.sort(score, seqs);
- if (sortLengthAscending)
+ if (sortBitScoreAscending)
{
setReverseOrder(align, seqs);
}
setOrder(align, seqs);
}
- sortLengthAscending = !sortLengthAscending;
+ sortBitScoreAscending = !sortBitScoreAscending;
}
/**
* object.
*
* @param align
- * sorts the given alignment object by group
+ * sorts the given alignment object by group
*/
public static void sortByGroup(AlignmentI align)
{
// MAINTAINS ORIGNAL SEQUENCE ORDER,
// ORDERS BY GROUP SIZE
- Vector groups = new Vector();
+ List<SequenceGroup> groups = new ArrayList<>();
- if (groups.hashCode() != lastGroupHash)
+ AlignmentSorter as = getInstance();
+
+ if (groups.hashCode() != as.lastGroupHash)
{
- sortGroupAscending = true;
- lastGroupHash = groups.hashCode();
+ as.sortGroupAscending = true;
+ as.lastGroupHash = groups.hashCode();
}
else
{
- sortGroupAscending = !sortGroupAscending;
+ as.sortGroupAscending = !as.sortGroupAscending;
}
// SORTS GROUPS BY SIZE
// ////////////////////
- for (int i = 0; i < align.getGroups().size(); i++)
+ for (SequenceGroup sg : align.getGroups())
{
- SequenceGroup sg = (SequenceGroup) align.getGroups().elementAt(i);
-
for (int j = 0; j < groups.size(); j++)
{
- SequenceGroup sg2 = (SequenceGroup) groups.elementAt(j);
+ SequenceGroup sg2 = groups.get(j);
if (sg.getSize() > sg2.getSize())
{
- groups.insertElementAt(sg, j);
+ groups.add(j, sg);
break;
}
if (!groups.contains(sg))
{
- groups.addElement(sg);
+ groups.add(sg);
}
}
// NOW ADD SEQUENCES MAINTAINING ALIGNMENT ORDER
// /////////////////////////////////////////////
- Vector seqs = new Vector();
+ List<SequenceI> tmp = new ArrayList<>();
for (int i = 0; i < groups.size(); i++)
{
- SequenceGroup sg = (SequenceGroup) groups.elementAt(i);
+ SequenceGroup sg = groups.get(i);
SequenceI[] orderedseqs = sg.getSequencesInOrder(align);
for (int j = 0; j < orderedseqs.length; j++)
{
- seqs.addElement(orderedseqs[j]);
- }
- }
-
- if (sortGroupAscending)
- {
- setOrder(align, seqs);
- }
- else
- {
- setReverseOrder(align,
- vectorSubsetToArray(seqs, align.getSequences()));
- }
- }
-
- /**
- * Converts Vector to array. java 1.18 does not have Vector.toArray()
- *
- * @param tmp
- * Vector of SequenceI objects
- *
- * @return array of Sequence[]
- */
- private static SequenceI[] vectorToArray(Vector tmp)
- {
- SequenceI[] seqs = new SequenceI[tmp.size()];
-
- for (int i = 0; i < tmp.size(); i++)
- {
- seqs[i] = (SequenceI) tmp.elementAt(i);
- }
-
- return seqs;
- }
-
- /**
- * DOCUMENT ME!
- *
- * @param tmp
- * DOCUMENT ME!
- * @param mask
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
- */
- private static SequenceI[] vectorSubsetToArray(Vector tmp, Vector mask)
- {
- Vector seqs = new Vector();
- int i;
- boolean[] tmask = new boolean[mask.size()];
-
- for (i = 0; i < mask.size(); i++)
- {
- tmask[i] = true;
- }
-
- for (i = 0; i < tmp.size(); i++)
- {
- Object sq = tmp.elementAt(i);
-
- if (mask.contains(sq) && tmask[mask.indexOf(sq)])
- {
- tmask[mask.indexOf(sq)] = false;
- seqs.addElement(sq);
- }
- }
-
- for (i = 0; i < tmask.length; i++)
- {
- if (tmask[i])
- {
- seqs.addElement(mask.elementAt(i));
+ tmp.add(orderedseqs[j]);
}
}
-
- return vectorToArray(seqs);
+ set(align, tmp, as.sortGroupAscending);
}
/**
* Sorts by a given AlignmentOrder object
*
* @param align
- * Alignment to order
+ * Alignment to order
* @param order
- * specified order for alignment
+ * specified order for alignment
*/
public static void sortBy(AlignmentI align, AlignmentOrder order)
{
// Get an ordered vector of sequences which may also be present in align
- Vector tmp = order.getOrder();
+ List<SequenceI> tmp = order.getOrder();
- if (lastOrder == order)
- {
- sortOrderAscending = !sortOrderAscending;
- }
- else
- {
- sortOrderAscending = true;
- }
+ AlignmentSorter as = getInstance();
- if (sortOrderAscending)
+ if (as.lastOrder == order)
{
- setOrder(align, tmp);
+ as.sortOrderAscending = !as.sortOrderAscending;
}
else
{
- setReverseOrder(align, vectorSubsetToArray(tmp, align.getSequences()));
+ as.sortOrderAscending = true;
}
+ set(align, tmp, as.sortOrderAscending);
}
/**
* DOCUMENT ME!
*
* @param align
- * alignment to order
+ * alignment to order
* @param tree
- * tree which has
+ * tree which has
*
* @return DOCUMENT ME!
*/
- private static Vector getOrderByTree(AlignmentI align, NJTree tree)
+ private static List<SequenceI> getOrderByTree(AlignmentI align,
+ TreeModel tree)
{
int nSeq = align.getHeight();
- Vector tmp = new Vector();
+ List<SequenceI> tmp = new ArrayList<>();
tmp = _sortByTree(tree.getTopNode(), tmp, align.getSequences());
// TODO: JBPNote - decide if this is always an error
// (eg. not when a tree is associated to another alignment which has more
// sequences)
- if (tmp.size() < nSeq)
+ if (tmp.size() != nSeq)
{
addStrays(align, tmp);
}
if (tmp.size() != nSeq)
{
- System.err.println("ERROR: tmp.size()=" + tmp.size() + " != nseq="
- + nSeq + " in getOrderByTree");
+ System.err.println("WARNING: tmp.size()=" + tmp.size() + " != nseq="
+ + nSeq
+ + " in getOrderByTree - tree contains sequences not in alignment");
}
}
* Sorts the alignment by a given tree
*
* @param align
- * alignment to order
+ * alignment to order
* @param tree
- * tree which has
+ * tree which has
*/
- public static void sortByTree(AlignmentI align, NJTree tree)
+ public static void sortByTree(AlignmentI align, TreeModel tree)
{
- Vector tmp = getOrderByTree(align, tree);
+ List<SequenceI> tmp = getOrderByTree(align, tree);
- // tmp should properly permute align with tree.
- if (lastTree != tree)
- {
- sortTreeAscending = true;
- lastTree = tree;
- }
- else
- {
- sortTreeAscending = !sortTreeAscending;
- }
+ AlignmentSorter as = getInstance();
- if (sortTreeAscending)
+ // tmp should properly permute align with tree.
+ if (as.lastTree != tree)
{
- setOrder(align, tmp);
+ as.sortTreeAscending = true;
+ as.lastTree = tree;
}
else
{
- setReverseOrder(align, vectorSubsetToArray(tmp, align.getSequences()));
+ as.sortTreeAscending = !as.sortTreeAscending;
}
+ set(align, tmp, as.sortTreeAscending);
}
/**
* DOCUMENT ME!
*
* @param align
- * DOCUMENT ME!
- * @param seqs
- * DOCUMENT ME!
+ * DOCUMENT ME!
+ * @param tmp
+ * DOCUMENT ME!
*/
- private static void addStrays(AlignmentI align, Vector seqs)
+ private static void addStrays(AlignmentI align, List<SequenceI> tmp)
{
int nSeq = align.getHeight();
for (int i = 0; i < nSeq; i++)
{
- if (!seqs.contains(align.getSequenceAt(i)))
+ if (!tmp.contains(align.getSequenceAt(i)))
{
- seqs.addElement(align.getSequenceAt(i));
+ tmp.add(align.getSequenceAt(i));
}
}
- if (nSeq != seqs.size())
+ if (nSeq != tmp.size())
{
System.err
.println("ERROR: Size still not right even after addStrays");
* DOCUMENT ME!
*
* @param node
- * DOCUMENT ME!
+ * DOCUMENT ME!
* @param tmp
- * DOCUMENT ME!
+ * DOCUMENT ME!
* @param seqset
- * DOCUMENT ME!
+ * DOCUMENT ME!
*
* @return DOCUMENT ME!
*/
- private static Vector _sortByTree(SequenceNode node, Vector tmp,
- Vector seqset)
+ private static List<SequenceI> _sortByTree(SequenceNode node,
+ List<SequenceI> tmp, List<SequenceI> seqset)
{
if (node == null)
{
{
if (node.element() instanceof SequenceI)
{
- if (!tmp.contains(node.element()))
+ if (!tmp.contains(node.element())) // && (seqset==null ||
+ // seqset.size()==0 ||
+ // seqset.contains(tmp)))
{
- tmp.addElement((SequenceI) node.element());
+ tmp.add((SequenceI) node.element());
}
}
}
for (int i = 0; i < alignment.length; i++)
{
- ids[i] = (new Float(alignment[i].getName().substring(8)))
+ ids[i] = (Float.valueOf(alignment[i].getName().substring(8)))
.floatValue();
}
* particular scoreLabel. Or reverse if same label was used previously
*
* @param scoreLabel
- * exact label for sequence associated AlignmentAnnotation
- * scores to use for sorting.
+ * exact label for sequence associated AlignmentAnnotation scores to
+ * use for sorting.
* @param alignment
- * sequences to be sorted
+ * sequences to be sorted
*/
public static void sortByAnnotationScore(String scoreLabel,
AlignmentI alignment)
{
SequenceI[] seqs = alignment.getSequencesArray();
boolean[] hasScore = new boolean[seqs.length]; // per sequence score
- // presence
+ // presence
int hasScores = 0; // number of scores present on set
double[] scores = new double[seqs.length];
double min = 0, max = 0;
hasScores++;
hasScore[i] = true;
scores[i] = scoreAnn[0].getScore(); // take the first instance of this
- // score.
+ // score.
if (hasScores == 1)
{
max = min = scores[i];
{
if (!hasScore[i])
{
- scores[i] = (max + i+1.0);
+ scores[i] = (max + i + 1.0);
}
}
}
jalview.util.QuickSort.sort(scores, seqs);
- if (lastSortByScore != scoreLabel)
+
+ AlignmentSorter as = getInstance();
+
+ if (as.lastSortByAnnotation != scoreLabel)
{
- lastSortByScore = scoreLabel;
+ as.lastSortByAnnotation = scoreLabel;
setOrder(alignment, seqs);
}
else
setReverseOrder(alignment, seqs);
}
}
+
/**
- * types of feature ordering:
- * Sort by score : average score - or total score - over all features in region
- * Sort by feature label text: (or if null - feature type text) - numerical or alphabetical
- * Sort by feature density: based on counts - ignoring individual text or scores for each feature
- */
- public static String FEATURE_SCORE="average_score";
- public static String FEATURE_LABEL="text";
- public static String FEATURE_DENSITY="density";
-
- /**
- * sort the alignment using the features on each sequence found between start and stop with the given featureLabel (and optional group qualifier)
- * @param featureLabel (may not be null)
- * @param groupLabel (may be null)
- * @param start (-1 to include non-positional features)
- * @param stop (-1 to only sort on non-positional features)
- * @param alignment - aligned sequences containing features
- * @param method - one of the string constants FEATURE_SCORE, FEATURE_LABEL, FEATURE_DENSITY
+ * Sort sequences by feature score or density, optionally restricted by
+ * feature types, feature groups, or alignment start/end positions.
+ * <p>
+ * If the sort is repeated for the same combination of types and groups, sort
+ * order is reversed.
+ *
+ * @param featureTypes
+ * a list of feature types to include (or null for all)
+ * @param groups
+ * a list of feature groups to include (or null for all)
+ * @param startCol
+ * start column position to include (base zero)
+ * @param endCol
+ * end column position to include (base zero)
+ * @param alignment
+ * the alignment to be sorted
+ * @param method
+ * either "average_score" or "density" ("text" not yet implemented)
*/
- public static void sortByFeature(String featureLabel, String groupLabel, int start, int stop,
+ public static void sortByFeature(List<String> featureTypes,
+ List<String> groups, final int startCol, final int endCol,
AlignmentI alignment, String method)
{
- sortByFeature(featureLabel==null ? null : new String[] {featureLabel},
- groupLabel==null ? null : new String[] {groupLabel}, start, stop, alignment, method);
- }
- private static boolean containsIgnoreCase(final String lab, final String[] labs)
- {
- if (labs==null)
- {
- return true;
- }
- if (lab==null)
- {
- return false;
- }
- for (int q=0;q<labs.length;q++)
- {
- if (labs[q]!=null && lab.equalsIgnoreCase(labs[q]))
- {
- return true;
- }
- }
- return false;
- }
- public static void sortByFeature(String[] featureLabels, String[] groupLabels, int start, int stop,
- AlignmentI alignment, String method)
- {
- if (method!=FEATURE_SCORE && method!=FEATURE_LABEL && method!=FEATURE_DENSITY)
- {
- throw new Error("Implementation Error - sortByFeature method must be one of FEATURE_SCORE, FEATURE_LABEL or FEATURE_DENSITY.");
- }
- boolean ignoreScore=method!=FEATURE_SCORE;
- StringBuffer scoreLabel = new StringBuffer();
- scoreLabel.append(start+stop+method);
- // This doesn't work yet - we'd like to have a canonical ordering that can be preserved from call to call
- for (int i=0;featureLabels!=null && i<featureLabels.length; i++)
- {
- scoreLabel.append(featureLabels[i]==null ? "null" : featureLabels[i]);
- }
- for (int i=0;groupLabels!=null && i<groupLabels.length; i++)
+ if (method != FEATURE_SCORE && method != FEATURE_LABEL
+ && method != FEATURE_DENSITY)
{
- scoreLabel.append(groupLabels[i]==null ? "null" : groupLabels[i]);
+ String msg = String.format(
+ "Implementation Error - sortByFeature method must be either '%s' or '%s'",
+ FEATURE_SCORE, FEATURE_DENSITY);
+ System.err.println(msg);
+ return;
}
+
+ flipFeatureSortIfUnchanged(method, featureTypes, groups, startCol,
+ endCol);
+
SequenceI[] seqs = alignment.getSequencesArray();
-
+
boolean[] hasScore = new boolean[seqs.length]; // per sequence score
- // presence
+ // presence
int hasScores = 0; // number of scores present on set
double[] scores = new double[seqs.length];
int[] seqScores = new int[seqs.length];
- Object[] feats = new Object[seqs.length];
- double min = 0, max = 0;
+ Object[][] feats = new Object[seqs.length][];
+ double min = 0d;
+ double max = 0d;
+
for (int i = 0; i < seqs.length; i++)
{
- SequenceFeature[] sf = seqs[i].getSequenceFeatures();
- if (sf==null && seqs[i].getDatasetSequence()!=null)
- {
- sf = seqs[i].getDatasetSequence().getSequenceFeatures();
- }
- if (sf==null)
+ /*
+ * get sequence residues overlapping column region
+ * and features for residue positions and specified types
+ */
+ String[] types = featureTypes == null ? null
+ : featureTypes.toArray(new String[featureTypes.size()]);
+ List<SequenceFeature> sfs = seqs[i].findFeatures(startCol + 1,
+ endCol + 1, types);
+
+ seqScores[i] = 0;
+ scores[i] = 0.0;
+
+ Iterator<SequenceFeature> it = sfs.listIterator();
+ while (it.hasNext())
{
- sf = new SequenceFeature[0];
- } else {
- SequenceFeature[] tmp = new SequenceFeature[sf.length];
- for (int s=0; s<tmp.length;s++)
+ SequenceFeature sf = it.next();
+
+ /*
+ * accept all features with null or empty group, otherwise
+ * check group is one of the currently visible groups
+ */
+ String featureGroup = sf.getFeatureGroup();
+ if (groups != null && featureGroup != null
+ && !"".equals(featureGroup)
+ && !groups.contains(featureGroup))
{
- tmp[s] = sf[s];
+ it.remove();
}
- sf = tmp;
- }
- int sstart = (start==-1) ? start : seqs[i].findPosition(start);
- int sstop = (stop==-1) ? stop : seqs[i].findPosition(stop);
- seqScores[i]=0;
- scores[i]=0.0;
- int n=sf.length;
- for (int f=0;f<sf.length;f++)
- {
- // filter for selection criteria
- if (
- // ignore features outwith alignment start-stop positions.
- (sf[f].end < sstart || sf[f].begin > sstop)
- ||
- // or ignore based on selection criteria
- (featureLabels != null && !AlignmentSorter.containsIgnoreCase(sf[f].type, featureLabels))
- || (groupLabels != null
- // problem here: we cannot eliminate null feature group features
- && (sf[f].getFeatureGroup() != null
- && !AlignmentSorter.containsIgnoreCase(sf[f].getFeatureGroup(), groupLabels))))
+ else
{
- // forget about this feature
- sf[f] = null;
- n--;
- } else {
- // or, also take a look at the scores if necessary.
- if (!ignoreScore && sf[f].getScore()!=Float.NaN)
+ float score = sf.getScore();
+ if (FEATURE_SCORE.equals(method) && !Float.isNaN(score))
{
- if (seqScores[i]==0)
+ if (seqScores[i] == 0)
{
hasScores++;
}
seqScores[i]++;
hasScore[i] = true;
- scores[i] += sf[f].getScore(); // take the first instance of this
- // score.
+ scores[i] += score;
+ // take the first instance of this score // ??
}
}
}
- SequenceFeature[] fs;
- feats[i] = fs = new SequenceFeature[n];
- if (n>0)
+
+ feats[i] = sfs.toArray(new SequenceFeature[sfs.size()]);
+ if (!sfs.isEmpty())
{
- n=0;
- for (int f=0;f<sf.length;f++)
- {
- if (sf[f]!=null)
- {
- ((SequenceFeature[]) feats[i])[n++] = sf[f];
- }
- }
- if (method==FEATURE_LABEL)
+ if (method == FEATURE_LABEL)
{
- // order the labels by alphabet
- String[] labs = new String[fs.length];
- for (int l=0;l<labs.length; l++)
+ // order the labels by alphabet (not yet implemented)
+ String[] labs = new String[sfs.size()];
+ for (int l = 0; l < sfs.size(); l++)
{
- labs[l] = (fs[l].getDescription()!=null ? fs[l].getDescription() : fs[l].getType());
+ SequenceFeature sf = sfs.get(l);
+ String description = sf.getDescription();
+ labs[l] = (description != null ? description : sf.getType());
}
- jalview.util.QuickSort.sort(labs, ((Object[]) feats[i]));
+ QuickSort.sort(labs, feats[i]);
}
}
if (hasScore[i])
- {
+ {
// compute average score
- scores[i]/=seqScores[i];
+ scores[i] /= seqScores[i];
// update the score bounds.
if (hasScores == 1)
{
- max = min = scores[i];
+ min = scores[i];
+ max = min;
}
else
{
- if (max < scores[i])
+ max = Math.max(max, scores[i]);
+ min = Math.min(min, scores[i]);
+ }
+ }
+ }
+
+ boolean doSort = false;
+
+ if (FEATURE_SCORE.equals(method))
+ {
+ if (hasScores == 0)
+ {
+ return; // do nothing - no scores present to sort by.
+ }
+ // pad score matrix
+ if (hasScores < seqs.length)
+ {
+ for (int i = 0; i < seqs.length; i++)
+ {
+ if (!hasScore[i])
{
- max = scores[i];
+ scores[i] = (max + 1 + i);
}
- if (min > scores[i])
+ else
{
- min = scores[i];
+ // int nf = (feats[i] == null) ? 0
+ // : ((SequenceFeature[]) feats[i]).length;
+ // // System.err.println("Sorting on Score: seq " +
+ // seqs[i].getName()
+ // + " Feats: " + nf + " Score : " + scores[i]);
}
}
}
+ doSort = true;
}
-
- if (method==FEATURE_SCORE)
+ else if (FEATURE_DENSITY.equals(method))
{
- if (hasScores == 0)
+ for (int i = 0; i < seqs.length; i++)
+ {
+ int featureCount = feats[i] == null ? 0
+ : ((SequenceFeature[]) feats[i]).length;
+ scores[i] = featureCount;
+ // System.err.println("Sorting on Density: seq "+seqs[i].getName()+
+ // " Feats: "+featureCount+" Score : "+scores[i]);
+ }
+ doSort = true;
+ }
+ if (doSort)
{
- return; // do nothing - no scores present to sort by.
+ QuickSort.sortByDouble(scores, seqs,
+ getInstance().sortByFeatureAscending);
}
- // pad score matrix
- if (hasScores < seqs.length)
+ setOrder(alignment, seqs);
+ }
+
+ /**
+ * Builds a string hash of criteria for sorting, and if unchanged from last
+ * time, reverse the sort order
+ *
+ * @param method
+ * @param featureTypes
+ * @param groups
+ * @param startCol
+ * @param endCol
+ */
+ protected static void flipFeatureSortIfUnchanged(String method,
+ List<String> featureTypes, List<String> groups,
+ final int startCol, final int endCol)
+ {
+ StringBuilder sb = new StringBuilder(64);
+ sb.append(startCol).append(method).append(endCol);
+ if (featureTypes != null)
+ {
+ Collections.sort(featureTypes);
+ sb.append(featureTypes.toString());
+ }
+ if (groups != null)
+ {
+ Collections.sort(groups);
+ sb.append(groups.toString());
+ }
+ String scoreCriteria = sb.toString();
+
+ /*
+ * if resorting on the same criteria, toggle sort order
+ */
+ AlignmentSorter as = getInstance();
+ if (as.sortByFeatureCriteria == null
+ || !scoreCriteria.equals(as.sortByFeatureCriteria))
+ {
+ as.sortByFeatureAscending = true;
+ }
+ else
+ {
+ as.sortByFeatureAscending = !as.sortByFeatureAscending;
+ }
+ as.sortByFeatureCriteria = scoreCriteria;
+ }
+
+ /**
+ * Set the alignment's sequences list to contain the sequences from a
+ * temporary list, first adding all the elements from the tmp list, then adding all sequences in the alignment that
+ * are not in the list. Option to do the final sort either in order or in reverse order.
+ *
+ * @param align The alignment being sorted
+ * @param tmp
+ * the temporary sequence list
+ * @param ascending
+ * false for reversed order; only sequences already in
+ * the alignment will be used (which is actually already guaranteed
+ * by vectorSubsetToArray)
+ */
+ private static void set(AlignmentI align, List<SequenceI> tmp,
+ boolean ascending)
+ {
+ set(align, vectorSubsetToArray(align.getSequences(), tmp), ascending);
+ }
+
+ /**
+ * Set the alignment's sequences list to contain these sequences, either in
+ * this order or its reverse.
+ *
+ * @param align
+ * @param seqs
+ * the new sequence array
+ * @param ascending
+ * false for reversed order; if ascending, only sequences already in
+ * the alignment will be used; if descending, then a direct 1:1
+ * replacement is made
+ */
+ private static void set(AlignmentI align, SequenceI[] seqs,
+ boolean ascending)
+ {
+ if (ascending)
+ {
+ setOrder(align, seqs);
+ }
+ else
{
+ setReverseOrder(align, seqs);
+ }
+
+ }
+
+ /**
+ * Replace the alignment's sequences with values in an array, clearing the
+ * alignment's sequence list and filtering for sequences that are actually in
+ * the alignment already.
+ *
+ * @param align
+ * the Alignment
+ * @param seqs
+ * the array of replacement values, of any length
+ */
+ public static void setOrder(AlignmentI align, SequenceI[] seqs)
+ {
+ // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work
+ List<SequenceI> seqList = align.getSequences();
+ synchronized (seqList)
+ {
+ List<SequenceI> tmp = new ArrayList<>();
+
for (int i = 0; i < seqs.length; i++)
{
- if (!hasScore[i])
+ if (seqList.contains(seqs[i]))
{
- scores[i] = (max + i);
- } else {
- int nf=(feats[i]==null) ? 0 :((SequenceFeature[]) feats[i]).length;
- System.err.println("Sorting on Score: seq "+seqs[i].getName()+ " Feats: "+nf+" Score : "+scores[i]);
+ tmp.add(seqs[i]);
}
}
- }
- jalview.util.QuickSort.sort(scores, seqs);
+ seqList.clear();
+ // User may have hidden seqs, then clicked undo or redo
+ for (int i = 0; i < tmp.size(); i++)
+ {
+ seqList.add(tmp.get(i));
+ }
}
- else
- if (method==FEATURE_DENSITY)
+ }
+
+ /**
+ * Replace the alignment's sequences or a subset of those sequences with
+ * values in an array in reverse order. All sequences are replaced; no check
+ * is made that these sequences are in the alignment already.
+ *
+ * @param align
+ * the Alignment
+ * @param seqs
+ * the array of replacement values, length must be less than or equal
+ * to Alignment.sequences.size()
+ */
+ private static void setReverseOrder(AlignmentI align, SequenceI[] seqs)
+ {
+ int nSeq = seqs.length;
+
+ int len = (nSeq + (nSeq % 2)) / 2;
+ // int len = 0;
+ //
+ // if ((nSeq % 2) == 0)
+ // {
+ // len = nSeq / 2;
+ // }
+ // else
+ // {
+ // len = (nSeq + 1) / 2;
+ // }
+
+ // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work
+ List<SequenceI> seqList = align.getSequences();
+ synchronized (seqList)
+ {
+ for (int i = 0; i < len; i++)
{
-
- // break ties between equivalent numbers for adjacent sequences by adding 1/Nseq*i on the original order
- double fr = 0.9/(1.0*seqs.length);
- for (int i=0;i<seqs.length; i++)
- {
- double nf;
- scores[i] = (0.05+fr*i)+(nf=((feats[i]==null) ? 0.0 :1.0*((SequenceFeature[]) feats[i]).length));
- System.err.println("Sorting on Density: seq "+seqs[i].getName()+ " Feats: "+nf+" Score : "+scores[i]);
- }
- jalview.util.QuickSort.sort(scores, seqs);
+ // SequenceI tmp = seqs[i];
+ seqList.set(i, seqs[nSeq - i - 1]);
+ seqList.set(nSeq - i - 1, seqs[i]);
}
- else {
- if (method==FEATURE_LABEL)
- {
- throw new Error("Not yet implemented.");
- }
+ }
+ }
+
+ /**
+ * Create and array of reordered sequences in order first from tmp that are
+ * present in seqList already, then, after that, any remaining sequences in
+ * seqList not in tmp. Any sequences in tmp that are not in seqList already
+ * are discarded.
+ *
+ * @param seqList
+ * thread safe collection of sequences originally in the alignment
+ * @param tmp
+ * thread safe collection of sequences or subsequences possibly in
+ * seqList
+ *
+ * @return intersect(tmp,seqList)+intersect(complement(tmp),seqList)
+ */
+ private static SequenceI[] vectorSubsetToArray(List<SequenceI> seqList,
+ List<SequenceI> tmp)
+ {
+ ArrayList<SequenceI> seqs = new ArrayList<>();
+ int n = seqList.size();
+ BitSet bs = new BitSet(n);
+ bs.set(0, n);
+ for (int i = 0, nt = tmp.size(); i < nt; i++)
+ {
+ SequenceI sq = tmp.get(i);
+ int idx = seqList.indexOf(sq);
+ if (idx >= 0 && bs.get(idx))
+ {
+ seqs.add(sq);
+ bs.clear(idx);
}
- if (lastSortByFeatureScore ==null || scoreLabel.equals(lastSortByFeatureScore))
- {
- setOrder(alignment, seqs);
}
- else
+
+ for (int i = bs.nextSetBit(0); i >= 0; i = bs.nextSetBit(i + 1))
{
- setReverseOrder(alignment, seqs);
+ seqs.add(seqList.get(i));
}
- lastSortByFeatureScore = scoreLabel.toString();
+
+ return seqs.toArray(new SequenceI[seqs.size()]);
}
}
git://source.jalview.org / jalview.git / blobdiff