-/*\r
-* Jalview - A Sequence Alignment Editor and Viewer\r
-* Copyright (C) 2006 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle\r
-*\r
-* This program is free software; you can redistribute it and/or\r
-* modify it under the terms of the GNU General Public License\r
-* as published by the Free Software Foundation; either version 2\r
-* of the License, or (at your option) any later version.\r
-*\r
-* This program is distributed in the hope that it will be useful,\r
-* but WITHOUT ANY WARRANTY; without even the implied warranty of\r
-* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\r
-* GNU General Public License for more details.\r
-*\r
-* You should have received a copy of the GNU General Public License\r
-* along with this program; if not, write to the Free Software\r
-* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA\r
-*/\r
-package jalview.analysis;\r
-\r
-import jalview.datamodel.*;\r
-\r
-import jalview.util.*;\r
-\r
-import java.util.*;\r
-\r
-\r
-/** Data structure to hold and manipulate a multiple sequence alignment\r
- */\r
-public class AlignmentSorter\r
-{\r
- static boolean sortIdAscending = true;\r
- static int lastGroupHash = 0;\r
- static boolean sortGroupAscending = true;\r
- static AlignmentOrder lastOrder = null;\r
- static boolean sortOrderAscending = true;\r
- static NJTree lastTree = null;\r
- static boolean sortTreeAscending = true;\r
-\r
- /**\r
- * Sort by Percentage Identity\r
- *\r
- * @param align AlignmentI\r
- * @param s SequenceI\r
- */\r
- public static void sortByPID(AlignmentI align, SequenceI s)\r
- {\r
- int nSeq = align.getHeight();\r
-\r
- float[] scores = new float[nSeq];\r
- SequenceI[] seqs = new SequenceI[nSeq];\r
-\r
- for (int i = 0; i < nSeq; i++)\r
- {\r
- scores[i] = Comparison.PID(align.getSequenceAt(i).getSequenceAsString(),\r
- s.getSequenceAsString());\r
- seqs[i] = align.getSequenceAt(i);\r
- }\r
-\r
- QuickSort.sort(scores, 0, scores.length - 1, seqs);\r
-\r
- setReverseOrder(align, seqs);\r
- }\r
-\r
- /**\r
- * Reverse the order of the sort\r
- *\r
- * @param align DOCUMENT ME!\r
- * @param seqs DOCUMENT ME!\r
- */\r
- private static void setReverseOrder(AlignmentI align, SequenceI[] seqs)\r
- {\r
- int nSeq = seqs.length;\r
-\r
- int len = 0;\r
-\r
- if ((nSeq % 2) == 0)\r
- {\r
- len = nSeq / 2;\r
- }\r
- else\r
- {\r
- len = (nSeq + 1) / 2;\r
- }\r
-\r
- // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work\r
- for (int i = 0; i < len; i++)\r
- {\r
- //SequenceI tmp = seqs[i];\r
- align.getSequences().setElementAt(seqs[nSeq - i - 1], i);\r
- align.getSequences().setElementAt(seqs[i], nSeq - i - 1);\r
- }\r
- }\r
-\r
- /**\r
- * Sets the Alignment object with the given sequences\r
- *\r
- * @param align Alignment object to be updated\r
- * @param tmp sequences as a vector\r
- */\r
- private static void setOrder(AlignmentI align, Vector tmp)\r
- {\r
- setOrder(align, vectorSubsetToArray(tmp, align.getSequences()));\r
- }\r
-\r
- /**\r
- * Sets the Alignment object with the given sequences\r
- *\r
- * @param align DOCUMENT ME!\r
- * @param seqs sequences as an array\r
- */\r
- public static void setOrder(AlignmentI align, SequenceI[] seqs)\r
- {\r
- // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work\r
- Vector algn = align.getSequences();\r
- Vector tmp = new Vector();\r
-\r
- for (int i = 0; i < seqs.length; i++)\r
- {\r
- if(algn.contains(seqs[i]))\r
- tmp.addElement(seqs[i]);\r
- }\r
-\r
- algn.removeAllElements();\r
- //User may have hidden seqs, then clicked undo or redo\r
- for (int i = 0; i < tmp.size(); i++)\r
- {\r
- algn.addElement(tmp.elementAt(i));\r
- }\r
-\r
-\r
-\r
-\r
- }\r
-\r
- /**\r
- * Sorts by ID. Numbers are sorted before letters.\r
- *\r
- * @param align The alignment object to sort\r
- */\r
- public static void sortByID(AlignmentI align)\r
- {\r
- int nSeq = align.getHeight();\r
-\r
- String[] ids = new String[nSeq];\r
- SequenceI[] seqs = new SequenceI[nSeq];\r
-\r
- for (int i = 0; i < nSeq; i++)\r
- {\r
- ids[i] = align.getSequenceAt(i).getName();\r
- seqs[i] = align.getSequenceAt(i);\r
- }\r
-\r
- QuickSort.sort(ids, seqs);\r
-\r
- if (sortIdAscending)\r
- {\r
- setReverseOrder(align, seqs);\r
- }\r
- else\r
- {\r
- setOrder(align, seqs);\r
- }\r
-\r
- sortIdAscending = !sortIdAscending;\r
- }\r
-\r
- /**\r
- * Sorts the alignment by size of group.\r
- * <br>Maintains the order of sequences in each group\r
- * by order in given alignment object.\r
- *\r
- * @param align sorts the given alignment object by group\r
- */\r
- public static void sortByGroup(AlignmentI align)\r
- {\r
- //MAINTAINS ORIGNAL SEQUENCE ORDER,\r
- //ORDERS BY GROUP SIZE\r
- Vector groups = new Vector();\r
-\r
- if (groups.hashCode() != lastGroupHash)\r
- {\r
- sortGroupAscending = true;\r
- lastGroupHash = groups.hashCode();\r
- }\r
- else\r
- {\r
- sortGroupAscending = !sortGroupAscending;\r
- }\r
-\r
- //SORTS GROUPS BY SIZE\r
- //////////////////////\r
- for (int i = 0; i < align.getGroups().size(); i++)\r
- {\r
- SequenceGroup sg = (SequenceGroup) align.getGroups().elementAt(i);\r
-\r
- for (int j = 0; j < groups.size(); j++)\r
- {\r
- SequenceGroup sg2 = (SequenceGroup) groups.elementAt(j);\r
-\r
- if (sg.getSize() > sg2.getSize())\r
- {\r
- groups.insertElementAt(sg, j);\r
-\r
- break;\r
- }\r
- }\r
-\r
- if (!groups.contains(sg))\r
- {\r
- groups.addElement(sg);\r
- }\r
- }\r
-\r
- //NOW ADD SEQUENCES MAINTAINING ALIGNMENT ORDER\r
- ///////////////////////////////////////////////\r
- Vector seqs = new Vector();\r
-\r
- for (int i = 0; i < groups.size(); i++)\r
- {\r
- SequenceGroup sg = (SequenceGroup) groups.elementAt(i);\r
- SequenceI[] orderedseqs = sg.getSequencesInOrder(align);\r
-\r
- for (int j = 0; j < orderedseqs.length; j++)\r
- {\r
- seqs.addElement(orderedseqs[j]);\r
- }\r
- }\r
-\r
- if (sortGroupAscending)\r
- {\r
- setOrder(align, seqs);\r
- }\r
- else\r
- {\r
- setReverseOrder(align,\r
- vectorSubsetToArray(seqs, align.getSequences()));\r
- }\r
- }\r
-\r
- /**\r
- * Converts Vector to array.\r
- * java 1.18 does not have Vector.toArray()\r
- *\r
- * @param tmp Vector of SequenceI objects\r
- *\r
- * @return array of Sequence[]\r
- */\r
- private static SequenceI[] vectorToArray(Vector tmp)\r
- {\r
- SequenceI[] seqs = new SequenceI[tmp.size()];\r
-\r
- for (int i = 0; i < tmp.size(); i++)\r
- {\r
- seqs[i] = (SequenceI) tmp.elementAt(i);\r
- }\r
-\r
- return seqs;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param tmp DOCUMENT ME!\r
- * @param mask DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- private static SequenceI[] vectorSubsetToArray(Vector tmp, Vector mask)\r
- {\r
- Vector seqs = new Vector();\r
- int i;\r
- boolean[] tmask = new boolean[mask.size()];\r
-\r
- for (i = 0; i < mask.size(); i++)\r
- tmask[i] = true;\r
-\r
- for (i = 0; i < tmp.size(); i++)\r
- {\r
- Object sq = tmp.elementAt(i);\r
-\r
- if (mask.contains(sq) && tmask[mask.indexOf(sq)])\r
- {\r
- tmask[mask.indexOf(sq)] = false;\r
- seqs.addElement(sq);\r
- }\r
- }\r
-\r
- for (i = 0; i < tmask.length; i++)\r
- if (tmask[i])\r
- {\r
- seqs.addElement(mask.elementAt(i));\r
- }\r
-\r
- return vectorToArray(seqs);\r
- }\r
-\r
- /**\r
- * Sorts by a given AlignmentOrder object\r
- *\r
- * @param align Alignment to order\r
- * @param order specified order for alignment\r
- */\r
- public static void sortBy(AlignmentI align, AlignmentOrder order)\r
- {\r
- // Get an ordered vector of sequences which may also be present in align\r
- Vector tmp = order.getOrder();\r
-\r
- if (lastOrder == order)\r
- {\r
- sortOrderAscending = !sortOrderAscending;\r
- }\r
- else\r
- {\r
- sortOrderAscending = true;\r
- }\r
-\r
- if (sortOrderAscending)\r
- {\r
- setOrder(align, tmp);\r
- }\r
- else\r
- {\r
- setReverseOrder(align,\r
- vectorSubsetToArray(tmp, align.getSequences()));\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param align alignment to order\r
- * @param tree tree which has\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- private static Vector getOrderByTree(AlignmentI align, NJTree tree)\r
- {\r
- int nSeq = align.getHeight();\r
-\r
- Vector tmp = new Vector();\r
-\r
- tmp = _sortByTree(tree.getTopNode(), tmp, align.getSequences());\r
-\r
- if (tmp.size() != nSeq)\r
- {\r
- // TODO: JBPNote - decide if this is always an error\r
- // (eg. not when a tree is associated to another alignment which has more\r
- // sequences)\r
- if (tmp.size() < nSeq)\r
- {\r
- addStrays(align, tmp);\r
- }\r
-\r
- if (tmp.size() != nSeq)\r
- {\r
- System.err.println("ERROR: tmp.size()=" + tmp.size() +\r
- " != nseq=" + nSeq + " in getOrderByTree");\r
- }\r
- }\r
-\r
- return tmp;\r
- }\r
-\r
- /**\r
- * Sorts the alignment by a given tree\r
- *\r
- * @param align alignment to order\r
- * @param tree tree which has\r
- */\r
- public static void sortByTree(AlignmentI align, NJTree tree)\r
- {\r
- Vector tmp = getOrderByTree(align, tree);\r
-\r
- // tmp should properly permute align with tree.\r
- if (lastTree != tree)\r
- {\r
- sortTreeAscending = true;\r
- lastTree = tree;\r
- }\r
- else\r
- {\r
- sortTreeAscending = !sortTreeAscending;\r
- }\r
-\r
- if (sortTreeAscending)\r
- {\r
- setOrder(align, tmp);\r
- }\r
- else\r
- {\r
- setReverseOrder(align,\r
- vectorSubsetToArray(tmp, align.getSequences()));\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param align DOCUMENT ME!\r
- * @param seqs DOCUMENT ME!\r
- */\r
- private static void addStrays(AlignmentI align, Vector seqs)\r
- {\r
- int nSeq = align.getHeight();\r
-\r
- for (int i = 0; i < nSeq; i++)\r
- {\r
- if (!seqs.contains(align.getSequenceAt(i)))\r
- {\r
- seqs.addElement(align.getSequenceAt(i));\r
- }\r
- }\r
-\r
- if (nSeq != seqs.size())\r
- {\r
- System.err.println(\r
- "ERROR: Size still not right even after addStrays");\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param tmp DOCUMENT ME!\r
- * @param seqset DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- private static Vector _sortByTree(SequenceNode node, Vector tmp,\r
- Vector seqset)\r
- {\r
- if (node == null)\r
- {\r
- return tmp;\r
- }\r
-\r
- SequenceNode left = (SequenceNode) node.left();\r
- SequenceNode right = (SequenceNode) node.right();\r
-\r
- if ((left == null) && (right == null))\r
- {\r
- if (!node.isPlaceholder() && (node.element() != null))\r
- {\r
- if (node.element() instanceof SequenceI)\r
- {\r
- if (!tmp.contains(node.element()))\r
- {\r
- tmp.addElement((SequenceI) node.element());\r
- }\r
- }\r
- }\r
-\r
- return tmp;\r
- }\r
- else\r
- {\r
- _sortByTree(left, tmp, seqset);\r
- _sortByTree(right, tmp, seqset);\r
- }\r
-\r
- return tmp;\r
- }\r
-\r
- // Ordering Objects\r
- // Alignment.sortBy(OrderObj) - sequence of sequence pointer refs in appropriate order\r
- //\r
-\r
- /**\r
- * recover the order of sequences given by the safe numbering scheme introducd\r
- * SeqsetUtils.uniquify.\r
- */\r
- public static void recoverOrder(SequenceI[] alignment)\r
- {\r
- float[] ids = new float[alignment.length];\r
-\r
- for (int i = 0; i < alignment.length; i++)\r
- ids[i] = (new Float(alignment[i].getName().substring(8))).floatValue();\r
-\r
- jalview.util.QuickSort.sort(ids, alignment);\r
- }\r
-}\r
+/*
+ * Jalview - A Sequence Alignment Editor and Viewer (Version 2.7)
+ * Copyright (C) 2011 J Procter, AM Waterhouse, J Engelhardt, LM Lui, G Barton, M Clamp, S Searle
+ *
+ * This file is part of Jalview.
+ *
+ * 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 Jalview. If not, see <http://www.gnu.org/licenses/>.
+ */
+package jalview.analysis;
+
+import java.util.*;
+
+import jalview.datamodel.*;
+import jalview.util.*;
+
+/**
+ * Routines for manipulating the order of a multiple sequence alignment TODO:
+ * this class retains some global states concerning sort-order which should be
+ * made attributes for the caller's alignment visualization. TODO: refactor to
+ * allow a subset of selected sequences to be sorted within the context of a
+ * whole alignment. Sort method template is: SequenceI[] tobesorted, [ input
+ * data mapping to each tobesorted element to use ], Alignment context of
+ * tobesorted that are to be re-ordered, boolean sortinplace, [special data - ie
+ * seuqence to be sorted w.r.t.]) sortinplace implies that the sorted vector
+ * resulting from applying the operation to tobesorted should be mapped back to
+ * the original positions in alignment. Otherwise, normal behaviour is to re
+ * order alignment so that tobesorted is sorted and grouped together starting
+ * 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
+{
+ /**
+ * todo: refactor searches to follow a basic pattern: (search property, last
+ * search state, current sort direction)
+ */
+ static boolean sortIdAscending = true;
+
+ static int lastGroupHash = 0;
+
+ static boolean sortGroupAscending = true;
+
+ static AlignmentOrder lastOrder = null;
+
+ static boolean sortOrderAscending = true;
+
+ static NJTree lastTree = null;
+
+ static boolean sortTreeAscending = true;
+
+ /**
+ * last Annotation Label used by sortByScore
+ */
+ private static String lastSortByScore;
+
+ private static boolean sortByScoreAscending = true;
+
+ /**
+ * compact representation of last arguments to SortByFeatureScore
+ */
+ private static String lastSortByFeatureScore;
+
+ private static boolean sortByFeatureScoreAscending = true;
+
+ private static boolean sortLengthAscending;
+
+ /**
+ * 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
+ *
+ * @param align
+ * AlignmentI
+ * @param s
+ * SequenceI
+ * @param tosort
+ * sequences from align that are to be sorted.
+ * @param start
+ * start column (0 for beginning
+ * @param end
+ */
+ public static void sortByPID(AlignmentI align, SequenceI s,
+ SequenceI[] tosort, int start, int end)
+ {
+ int nSeq = align.getHeight();
+
+ float[] scores = new float[nSeq];
+ SequenceI[] seqs = new SequenceI[nSeq];
+
+ for (int i = 0; i < nSeq; i++)
+ {
+ scores[i] = Comparison.PID(align.getSequenceAt(i)
+ .getSequenceAsString(), s.getSequenceAsString());
+ seqs[i] = align.getSequenceAt(i);
+ }
+
+ QuickSort.sort(scores, 0, scores.length - 1, seqs);
+
+ setReverseOrder(align, seqs);
+ }
+
+ /**
+ * Reverse the order of the sort
+ *
+ * @param align
+ * DOCUMENT ME!
+ * @param seqs
+ * DOCUMENT ME!
+ */
+ private static void setReverseOrder(AlignmentI align, SequenceI[] seqs)
+ {
+ int nSeq = seqs.length;
+
+ 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
+ for (int i = 0; i < len; i++)
+ {
+ // SequenceI tmp = seqs[i];
+ align.getSequences().setElementAt(seqs[nSeq - i - 1], i);
+ align.getSequences().setElementAt(seqs[i], nSeq - i - 1);
+ }
+ }
+
+ /**
+ * 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()));
+ }
+
+ /**
+ * Sets the Alignment object with the given sequences
+ *
+ * @param align
+ * DOCUMENT ME!
+ * @param seqs
+ * sequences as an array
+ */
+ public static void setOrder(AlignmentI align, SequenceI[] seqs)
+ {
+ // NOTE: DO NOT USE align.setSequenceAt() here - it will NOT work
+ Vector algn = align.getSequences();
+ Vector tmp = new Vector();
+
+ for (int i = 0; i < seqs.length; i++)
+ {
+ if (algn.contains(seqs[i]))
+ {
+ tmp.addElement(seqs[i]);
+ }
+ }
+
+ algn.removeAllElements();
+ // User may have hidden seqs, then clicked undo or redo
+ for (int i = 0; i < tmp.size(); i++)
+ {
+ algn.addElement(tmp.elementAt(i));
+ }
+
+ }
+
+ /**
+ * Sorts by ID. Numbers are sorted before letters.
+ *
+ * @param align
+ * The alignment object to sort
+ */
+ public static void sortByID(AlignmentI align)
+ {
+ int nSeq = align.getHeight();
+
+ String[] ids = new String[nSeq];
+ SequenceI[] seqs = new SequenceI[nSeq];
+
+ for (int i = 0; i < nSeq; i++)
+ {
+ ids[i] = align.getSequenceAt(i).getName();
+ seqs[i] = align.getSequenceAt(i);
+ }
+
+ QuickSort.sort(ids, seqs);
+
+ if (sortIdAscending)
+ {
+ setReverseOrder(align, seqs);
+ }
+ else
+ {
+ setOrder(align, seqs);
+ }
+
+ sortIdAscending = !sortIdAscending;
+ }
+
+ /**
+ * Sorts by sequence length
+ *
+ * @param align
+ * The alignment object to sort
+ */
+ public static void sortByLength(AlignmentI align)
+ {
+ int nSeq = align.getHeight();
+
+ float[] length = new float[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());
+ }
+
+ QuickSort.sort(length, seqs);
+
+ if (sortLengthAscending)
+ {
+ setReverseOrder(align, seqs);
+ }
+ else
+ {
+ setOrder(align, seqs);
+ }
+
+ sortLengthAscending = !sortLengthAscending;
+ }
+
+ /**
+ * Sorts the alignment by size of group. <br>
+ * Maintains the order of sequences in each group by order in given alignment
+ * object.
+ *
+ * @param align
+ * 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();
+
+ if (groups.hashCode() != lastGroupHash)
+ {
+ sortGroupAscending = true;
+ lastGroupHash = groups.hashCode();
+ }
+ else
+ {
+ sortGroupAscending = !sortGroupAscending;
+ }
+
+ // SORTS GROUPS BY SIZE
+ // ////////////////////
+ for (int i = 0; i < align.getGroups().size(); i++)
+ {
+ SequenceGroup sg = (SequenceGroup) align.getGroups().elementAt(i);
+
+ for (int j = 0; j < groups.size(); j++)
+ {
+ SequenceGroup sg2 = (SequenceGroup) groups.elementAt(j);
+
+ if (sg.getSize() > sg2.getSize())
+ {
+ groups.insertElementAt(sg, j);
+
+ break;
+ }
+ }
+
+ if (!groups.contains(sg))
+ {
+ groups.addElement(sg);
+ }
+ }
+
+ // NOW ADD SEQUENCES MAINTAINING ALIGNMENT ORDER
+ // /////////////////////////////////////////////
+ Vector seqs = new Vector();
+
+ for (int i = 0; i < groups.size(); i++)
+ {
+ SequenceGroup sg = (SequenceGroup) groups.elementAt(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, idx;
+ 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);
+ idx = mask.indexOf(sq);
+ if (idx > -1 && tmask[idx])
+ {
+ tmask[idx] = false;
+ seqs.addElement(sq);
+ }
+ }
+
+ for (i = 0; i < tmask.length; i++)
+ {
+ if (tmask[i])
+ {
+ seqs.addElement(mask.elementAt(i));
+ }
+ }
+
+ return vectorToArray(seqs);
+ }
+
+ /**
+ * Sorts by a given AlignmentOrder object
+ *
+ * @param align
+ * Alignment to order
+ * @param order
+ * 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();
+
+ if (lastOrder == order)
+ {
+ sortOrderAscending = !sortOrderAscending;
+ }
+ else
+ {
+ sortOrderAscending = true;
+ }
+
+ if (sortOrderAscending)
+ {
+ setOrder(align, tmp);
+ }
+ else
+ {
+ setReverseOrder(align, vectorSubsetToArray(tmp, align.getSequences()));
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param align
+ * alignment to order
+ * @param tree
+ * tree which has
+ *
+ * @return DOCUMENT ME!
+ */
+ private static Vector getOrderByTree(AlignmentI align, NJTree tree)
+ {
+ int nSeq = align.getHeight();
+
+ Vector tmp = new Vector();
+
+ tmp = _sortByTree(tree.getTopNode(), tmp, align.getSequences());
+
+ if (tmp.size() != nSeq)
+ {
+ // 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)
+ {
+ addStrays(align, tmp);
+ }
+
+ if (tmp.size() != nSeq)
+ {
+ System.err.println("WARNING: tmp.size()=" + tmp.size() + " != nseq="
+ + nSeq + " in getOrderByTree - tree contains sequences not in alignment");
+ }
+ }
+
+ return tmp;
+ }
+
+ /**
+ * Sorts the alignment by a given tree
+ *
+ * @param align
+ * alignment to order
+ * @param tree
+ * tree which has
+ */
+ public static void sortByTree(AlignmentI align, NJTree tree)
+ {
+ Vector tmp = getOrderByTree(align, tree);
+
+ // tmp should properly permute align with tree.
+ if (lastTree != tree)
+ {
+ sortTreeAscending = true;
+ lastTree = tree;
+ }
+ else
+ {
+ sortTreeAscending = !sortTreeAscending;
+ }
+
+ if (sortTreeAscending)
+ {
+ setOrder(align, tmp);
+ }
+ else
+ {
+ setReverseOrder(align, vectorSubsetToArray(tmp, align.getSequences()));
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param align
+ * DOCUMENT ME!
+ * @param seqs
+ * DOCUMENT ME!
+ */
+ private static void addStrays(AlignmentI align, Vector seqs)
+ {
+ int nSeq = align.getHeight();
+
+ for (int i = 0; i < nSeq; i++)
+ {
+ if (!seqs.contains(align.getSequenceAt(i)))
+ {
+ seqs.addElement(align.getSequenceAt(i));
+ }
+ }
+
+ if (nSeq != seqs.size())
+ {
+ System.err
+ .println("ERROR: Size still not right even after addStrays");
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ * @param tmp
+ * DOCUMENT ME!
+ * @param seqset
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ private static Vector _sortByTree(SequenceNode node, Vector tmp,
+ Vector seqset)
+ {
+ if (node == null)
+ {
+ return tmp;
+ }
+
+ SequenceNode left = (SequenceNode) node.left();
+ SequenceNode right = (SequenceNode) node.right();
+
+ if ((left == null) && (right == null))
+ {
+ if (!node.isPlaceholder() && (node.element() != null))
+ {
+ if (node.element() instanceof SequenceI)
+ {
+ if (!tmp.contains(node.element())) // && (seqset==null || seqset.size()==0 || seqset.contains(tmp)))
+ {
+ tmp.addElement((SequenceI) node.element());
+ }
+ }
+ }
+
+ return tmp;
+ }
+ else
+ {
+ _sortByTree(left, tmp, seqset);
+ _sortByTree(right, tmp, seqset);
+ }
+
+ return tmp;
+ }
+
+ // Ordering Objects
+ // Alignment.sortBy(OrderObj) - sequence of sequence pointer refs in
+ // appropriate order
+ //
+
+ /**
+ * recover the order of sequences given by the safe numbering scheme introducd
+ * SeqsetUtils.uniquify.
+ */
+ public static void recoverOrder(SequenceI[] alignment)
+ {
+ float[] ids = new float[alignment.length];
+
+ for (int i = 0; i < alignment.length; i++)
+ {
+ ids[i] = (new Float(alignment[i].getName().substring(8)))
+ .floatValue();
+ }
+
+ jalview.util.QuickSort.sort(ids, alignment);
+ }
+
+ /**
+ * Sort sequence in order of increasing score attribute for annotation with a
+ * particular scoreLabel. Or reverse if same label was used previously
+ *
+ * @param scoreLabel
+ * exact label for sequence associated AlignmentAnnotation scores to
+ * use for sorting.
+ * @param alignment
+ * 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
+ int hasScores = 0; // number of scores present on set
+ double[] scores = new double[seqs.length];
+ double min = 0, max = 0;
+ for (int i = 0; i < seqs.length; i++)
+ {
+ AlignmentAnnotation[] scoreAnn = seqs[i].getAnnotation(scoreLabel);
+ if (scoreAnn != null)
+ {
+ hasScores++;
+ hasScore[i] = true;
+ scores[i] = scoreAnn[0].getScore(); // take the first instance of this
+ // score.
+ if (hasScores == 1)
+ {
+ max = min = scores[i];
+ }
+ else
+ {
+ if (max < scores[i])
+ {
+ max = scores[i];
+ }
+ if (min > scores[i])
+ {
+ min = scores[i];
+ }
+ }
+ }
+ else
+ {
+ hasScore[i] = false;
+ }
+ }
+ if (hasScores == 0)
+ {
+ return; // do nothing - no scores present to sort by.
+ }
+ if (hasScores < seqs.length)
+ {
+ for (int i = 0; i < seqs.length; i++)
+ {
+ if (!hasScore[i])
+ {
+ scores[i] = (max + i + 1.0);
+ }
+ }
+ }
+
+ jalview.util.QuickSort.sort(scores, seqs);
+ if (lastSortByScore != scoreLabel)
+ {
+ lastSortByScore = 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
+ */
+ public static void sortByFeature(String featureLabel, String groupLabel,
+ int start, int stop, 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 quite 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++)
+ {
+ scoreLabel.append(groupLabels[i] == null ? "null" : groupLabels[i]);
+ }
+ SequenceI[] seqs = alignment.getSequencesArray();
+
+ boolean[] hasScore = new boolean[seqs.length]; // per sequence score
+ // 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;
+ 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)
+ {
+ sf = new SequenceFeature[0];
+ }
+ else
+ {
+ SequenceFeature[] tmp = new SequenceFeature[sf.length];
+ for (int s = 0; s < tmp.length; s++)
+ {
+ tmp[s] = sf[s];
+ }
+ 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))))
+ {
+ // 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)
+ {
+ if (seqScores[i] == 0)
+ {
+ hasScores++;
+ }
+ seqScores[i]++;
+ hasScore[i] = true;
+ scores[i] += sf[f].getScore(); // take the first instance of this
+ // score.
+ }
+ }
+ }
+ SequenceFeature[] fs;
+ feats[i] = fs = new SequenceFeature[n];
+ if (n > 0)
+ {
+ n = 0;
+ for (int f = 0; f < sf.length; f++)
+ {
+ if (sf[f] != null)
+ {
+ ((SequenceFeature[]) feats[i])[n++] = sf[f];
+ }
+ }
+ if (method == FEATURE_LABEL)
+ {
+ // order the labels by alphabet
+ String[] labs = new String[fs.length];
+ for (int l = 0; l < labs.length; l++)
+ {
+ labs[l] = (fs[l].getDescription() != null ? fs[l]
+ .getDescription() : fs[l].getType());
+ }
+ jalview.util.QuickSort.sort(labs, ((Object[]) feats[i]));
+ }
+ }
+ if (hasScore[i])
+ {
+ // compute average score
+ scores[i] /= seqScores[i];
+ // update the score bounds.
+ if (hasScores == 1)
+ {
+ max = min = scores[i];
+ }
+ else
+ {
+ if (max < scores[i])
+ {
+ max = scores[i];
+ }
+ if (min > scores[i])
+ {
+ min = scores[i];
+ }
+ }
+ }
+ }
+
+ if (method == FEATURE_SCORE)
+ {
+ 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])
+ {
+ scores[i] = (max + 1 + 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]);
+ }
+ }
+ }
+
+ jalview.util.QuickSort.sort(scores, seqs);
+ }
+ else if (method == FEATURE_DENSITY)
+ {
+
+ // 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);
+ }
+ else
+ {
+ if (method == FEATURE_LABEL)
+ {
+ throw new Error("Not yet implemented.");
+ }
+ }
+ if (lastSortByFeatureScore == null
+ || !scoreLabel.toString().equals(lastSortByFeatureScore))
+ {
+ sortByFeatureScoreAscending = true;
+ }
+ else
+ {
+ sortByFeatureScoreAscending = !sortByFeatureScoreAscending;
+ }
+ if (sortByFeatureScoreAscending)
+ {
+ setOrder(alignment, seqs);
+ }
+ else
+ {
+ setReverseOrder(alignment, seqs);
+ }
+ lastSortByFeatureScore = scoreLabel.toString();
+ }
+
+}
git://source.jalview.org / jalview.git / blobdiff