Added functional additional combinational filters for words

master
Luka 6 years ago
parent e2ce656fc5
commit c073e12f55

@ -6,14 +6,11 @@ import java.util.List;
import java.util.regex.Pattern;
import java.util.stream.Collectors;
import data.*;
import org.apache.commons.lang3.StringUtils;
import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import data.CalculateFor;
import data.Sentence;
import data.StatisticsNew;
import data.Word;
import gui.ValidationUtil;
public class Ngrams {
@ -45,9 +42,26 @@ public class Ngrams {
continue;
}
// generate proper MultipleHMKeys depending on filter data
String key = wordToString(ngramCandidate, stats.getFilter().getCalculateFor());
String lemma = "";
String wordType = "";
String msd = "";
for (CalculateFor otherKey : stats.getFilter().getMultipleKeys()){
if(otherKey.toString().equals("lema")){
lemma = wordToString(ngramCandidate, otherKey);
} else if(otherKey.toString().equals("besedna vrsta")){
wordType = wordToString(ngramCandidate, otherKey).substring(0, 1);
} else if(otherKey.toString().equals("oblikoskladenjska oznaka")){
msd = wordToString(ngramCandidate, otherKey);
}
}
MultipleHMKeys multipleKeys = new MultipleHMKeys(key, lemma, wordType, msd);
// UPDATE TAXONOMY HERE!!!
stats.updateTaxonomyResults(wordToString(ngramCandidate, stats.getFilter().getCalculateFor()), ngramCandidate.get(0).getTaxonomy());
stats.updateResults(wordToString(ngramCandidate, stats.getFilter().getCalculateFor()));
stats.updateTaxonomyResults(multipleKeys, ngramCandidate.get(0).getTaxonomy());
// stats.updateResults(wordToString(ngramCandidate, stats.getFilter().getCalculateFor()));
}
}
}
@ -129,7 +143,9 @@ public class Ngrams {
for (int i = 0; i < word.length() - stats.getFilter().getStringLength() + 1; i++) {
// TODO: locila?
stats.updateTaxonomyResults(word.substring(i, i + stats.getFilter().getStringLength()), taxonomy);
MultipleHMKeys multipleKeys = new MultipleHMKeys(word.substring(i, i + stats.getFilter().getStringLength()));
stats.updateTaxonomyResults(multipleKeys, taxonomy);
// stats.updateResults(wordToString(ngramCandidate, stats.getFilter().getCalculateFor()));

@ -24,7 +24,8 @@ public class Filter {
TAXONOMY,
MSD,
HAS_MSD,
SOLAR_FILTERS
SOLAR_FILTERS,
MULTIPLE_KEYS
}
public Filter() {
@ -141,4 +142,23 @@ public class Filter {
public HashMap<String, HashSet<String>> getSolarFilters() {
return (HashMap<String, HashSet<String>>) filter.get(SOLAR_FILTERS);
}
public void setMultipleKeys(ArrayList<String> keys) {
ArrayList<CalculateFor> newKeys = new ArrayList<>();
if (keys != null) {
for (String key : keys) {
newKeys.add(CalculateFor.factory(key));
}
}
filter.put(MULTIPLE_KEYS, newKeys);
}
public ArrayList<CalculateFor> getMultipleKeys() {
if (filter.containsKey(MULTIPLE_KEYS) && filter.get(MULTIPLE_KEYS) != null) {
return (ArrayList<CalculateFor>) filter.get(MULTIPLE_KEYS);
} else {
return new ArrayList<>();
}
}
}

@ -2,48 +2,54 @@ package data;
/*
Created for when words are sorted by multiple keys, i.e. not just lemmas but lemmas and msd simultaneously.
*/
final class MultipleHMKeys {
private final String key1, key2, key3;
public final class MultipleHMKeys {
private final String key, lemma, wordType, msd;
public MultipleHMKeys(String key1) {
this.key1 = key1;
this.key2 = null;
this.key3 = null;
public MultipleHMKeys(String key) {
this.key = key;
this.lemma = "";
this.wordType = "";
this.msd = "";
}
public MultipleHMKeys(String key1, String key2) {
this.key1 = key1;
this.key2 = key2;
this.key3 = null;
public MultipleHMKeys(String key, String lemma, String wordType, String msd) {
this.key = key;
this.lemma = lemma;
this.wordType = wordType;
this.msd = msd;
}
public MultipleHMKeys(String key1, String key2, String key3) {
this.key1 = key1;
this.key2 = key2;
this.key3 = key3;
public String getKey() {
return key;
}
public String getKey1() {
return key1;
public String getLemma() {
return lemma;
}
public String getKey2() {
return key2;
public String getWordType() {
return wordType;
}
public String getKey3() {
return key3;
public String getMsd() {
return msd;
}
@Override
public int hashCode() {
return key1.hashCode() ^ key2.hashCode() ^ key3.hashCode();
// if(key2 == null){
// return key1.hashCode();
// } else if (key3 == null){
// return key1.hashCode() ^ key2.hashCode();
// }
return key.hashCode() ^ lemma.hashCode() ^ wordType.hashCode() ^ msd.hashCode();
}
@Override
public boolean equals(Object obj) {
return (obj instanceof MultipleHMKeys) && ((MultipleHMKeys) obj).key1.equals(key1)
&& ((MultipleHMKeys) obj).key2.equals(key2)
&& ((MultipleHMKeys) obj).key3.equals(key3);
return (obj instanceof MultipleHMKeys) && ((MultipleHMKeys) obj).key.equals(key)
&& ((MultipleHMKeys) obj).lemma.equals(lemma)
&& ((MultipleHMKeys) obj).wordType.equals(wordType)
&& ((MultipleHMKeys) obj).msd.equals(msd);
}
}

@ -222,7 +222,7 @@ public class Statistics {
// return sortedM;
// }
private Map<String, Long> getSortedResult(Map<String, AtomicLong> map, int limit) {
private Map<MultipleHMKeys, Long> getSortedResult(Map<MultipleHMKeys, AtomicLong> map, int limit) {
return Util.sortByValue(Util.atomicInt2StringAndInt(map), limit);
}

@ -32,10 +32,10 @@ public class StatisticsNew {
private String resultTitle;
private Map<String, AtomicLong> result;
private Map<String, Map<String, AtomicLong>> taxonomyResult;
private Map<String, Map<MultipleHMKeys, AtomicLong>> taxonomyResult;
private Object[][] resultCustom; // for when calculating percentages that don't add up to 100%
private Map<String, ConcurrentHashMap<String, AtomicLong>> resultNestedSuffix;
private Map<String, ConcurrentHashMap<String, AtomicLong>> resultNestedPrefix;
private Map<String, ConcurrentHashMap<MultipleHMKeys, AtomicLong>> resultNestedSuffix;
private Map<String, ConcurrentHashMap<MultipleHMKeys, AtomicLong>> resultNestedPrefix;
private boolean useDB;
private RDB db;
private boolean analysisProducedResults;
@ -194,7 +194,7 @@ public class StatisticsNew {
}
public boolean saveResultToDisk(int... limit) throws UnsupportedEncodingException {
Set<Pair<String, Map<String, Long>>> stats = new HashSet<>();
Set<Pair<String, Map<MultipleHMKeys, Long>>> stats = new HashSet<>();
if (useDB) {
result = db.getDump();
@ -223,13 +223,14 @@ public class StatisticsNew {
}
Map<WordLevelType, Map<String, Map<String, Long>>> results = new HashMap<>();
if (!isEmpty(resultNestedSuffix)) {
results.put(WordLevelType.SUFFIX, sortNestedMap(resultNestedSuffix, Util.getValidInt(limit)));
}
if (!isEmpty(resultNestedPrefix)) {
results.put(WordLevelType.PREFIX, sortNestedMap(resultNestedPrefix, Util.getValidInt(limit)));
}
// UNCOMMENT!!!!!!
// if (!isEmpty(resultNestedSuffix)) {
// results.put(WordLevelType.SUFFIX, sortNestedMap(resultNestedSuffix, Util.getValidInt(limit)));
// }
//
// if (!isEmpty(resultNestedPrefix)) {
// results.put(WordLevelType.PREFIX, sortNestedMap(resultNestedPrefix, Util.getValidInt(limit)));
// }
// if no results and nothing to save, return false
if (!(results.size() > 0)) {
@ -266,8 +267,8 @@ public class StatisticsNew {
return true;
}
private Map<String, Map<String, Long>> sortNestedMap(Map<String, ConcurrentHashMap<String, AtomicLong>> nestedMap, int limit) {
Map<String, Map<String, Long>> sorted = new HashMap<>();
private Map<String, Map<MultipleHMKeys, Long>> sortNestedMap(Map<String, ConcurrentHashMap<MultipleHMKeys, AtomicLong>> nestedMap, int limit) {
Map<String, Map<MultipleHMKeys, Long>> sorted = new HashMap<>();
for (String s : nestedMap.keySet()) {
sorted.put(s, getSortedResult(nestedMap.get(s), Util.getValidInt(limit)));
@ -277,11 +278,11 @@ public class StatisticsNew {
}
private Map<String, Long> getSortedResult(Map<String, AtomicLong> map, int limit) {
private Map<MultipleHMKeys, Long> getSortedResult(Map<MultipleHMKeys, AtomicLong> map, int limit) {
return Util.sortByValue(Util.atomicInt2StringAndInt(map), limit);
}
public void updateTaxonomyResults(String o, List<String> taxonomy) {
public void updateTaxonomyResults(MultipleHMKeys o, List<String> taxonomy) {
for (String key : taxonomyResult.keySet()) {
// first word should have the same taxonomy as others
if (taxonomy.contains(key) || key.equals("Total")) {
@ -335,9 +336,11 @@ public class StatisticsNew {
}
public void updateResultsNestedSuffix(String key, String stringValue) {
MultipleHMKeys mkStringValue = new MultipleHMKeys(stringValue);
if (resultNestedSuffix.containsKey(key)) {
// if not in map
AtomicLong r = resultNestedSuffix.get(key).putIfAbsent(stringValue, new AtomicLong(1));
AtomicLong r = resultNestedSuffix.get(key).putIfAbsent(mkStringValue, new AtomicLong(1));
// else
if (r != null) {
@ -345,7 +348,7 @@ public class StatisticsNew {
}
} else {
resultNestedSuffix.putIfAbsent(key, new ConcurrentHashMap<>());
AtomicLong r = resultNestedSuffix.get(key).putIfAbsent(stringValue, new AtomicLong(1));
AtomicLong r = resultNestedSuffix.get(key).putIfAbsent(mkStringValue, new AtomicLong(1));
if (r != null) {
resultNestedSuffix.get(key).get(stringValue).incrementAndGet();
@ -354,9 +357,11 @@ public class StatisticsNew {
}
public void updateResultsNestedPrefix(String key, String stringValue) {
MultipleHMKeys mkStringValue = new MultipleHMKeys(stringValue);
if (resultNestedPrefix.containsKey(key)) {
// if not in map
AtomicLong r = resultNestedPrefix.get(key).putIfAbsent(stringValue, new AtomicLong(1));
AtomicLong r = resultNestedPrefix.get(key).putIfAbsent(mkStringValue, new AtomicLong(1));
// else
if (r != null) {
@ -364,7 +369,7 @@ public class StatisticsNew {
}
} else {
resultNestedPrefix.putIfAbsent(key, new ConcurrentHashMap<>());
AtomicLong r = resultNestedPrefix.get(key).putIfAbsent(stringValue, new AtomicLong(1));
AtomicLong r = resultNestedPrefix.get(key).putIfAbsent(mkStringValue, new AtomicLong(1));
if (r != null) {
resultNestedPrefix.get(key).get(stringValue).incrementAndGet();

@ -82,6 +82,7 @@ public class OneWordAnalysisTab {
private static final ObservableList<String> N_GRAM_COMPUTE_FOR_WORDS_ORTH = FXCollections.observableArrayList("različnica");
private static final ObservableList<String> alsoVisualizeItemsLemma = FXCollections.observableArrayList("besedna vrsta", "oblikoskladenjska oznaka");
private static final ObservableList<String> alsoVisualizeItemsDifferential = FXCollections.observableArrayList("lema", "besedna vrsta", "oblikoskladenjska oznaka");
private static final ObservableList<String> alsoVisualizeItemsEmpty = FXCollections.observableArrayList();
// TODO: pass observables for taxonomy based on header scan
// after header scan
@ -95,6 +96,37 @@ public class OneWordAnalysisTab {
// calculateForCB
calculateForCB.valueProperty().addListener((observable, oldValue, newValue) -> {
calculateFor = CalculateFor.factory(newValue);
if(newValue.equals("lema")){
alsoVisualizeCCB.getItems().removeAll();
alsoVisualizeCCB.getItems().setAll(alsoVisualizeItemsLemma);
alsoVisualizeCCB.getCheckModel().getCheckedItems().addListener((ListChangeListener<String>) c -> {
alsoVisualize = new ArrayList<>();
ObservableList<String> checkedItems = alsoVisualizeCCB.getCheckModel().getCheckedItems();
alsoVisualize.addAll(checkedItems);
logger.info(String.format("Selected also visualize items: %s", StringUtils.join(checkedItems, ",")));
});
alsoVisualizeCCB.getCheckModel().clearChecks();
} else if(newValue.equals("različnica")){
alsoVisualizeCCB.getItems().removeAll();
alsoVisualizeCCB.getItems().setAll(alsoVisualizeItemsDifferential);
alsoVisualizeCCB.getCheckModel().getCheckedItems().addListener((ListChangeListener<String>) c -> {
alsoVisualize = new ArrayList<>();
ObservableList<String> checkedItems = alsoVisualizeCCB.getCheckModel().getCheckedItems();
alsoVisualize.addAll(checkedItems);
logger.info(String.format("Selected also visualize items: %s", StringUtils.join(checkedItems, ",")));
});
alsoVisualizeCCB.getCheckModel().clearChecks();
} else {
alsoVisualizeCCB.getItems().removeAll();
alsoVisualizeCCB.getItems().setAll(alsoVisualizeItemsEmpty);
alsoVisualizeCCB.getCheckModel().getCheckedItems().addListener((ListChangeListener<String>) c -> {
alsoVisualize = new ArrayList<>();
ObservableList<String> checkedItems = alsoVisualizeCCB.getCheckModel().getCheckedItems();
alsoVisualize.addAll(checkedItems);
logger.info(String.format("Selected also visualize items: %s", StringUtils.join(checkedItems, ",")));
});
alsoVisualizeCCB.getCheckModel().clearChecks();
}
logger.info("calculateForCB:", calculateFor.toString());
});
@ -294,6 +326,7 @@ public class OneWordAnalysisTab {
filter.setIsCvv(false);
filter.setSolarFilters(solarFiltersMap);
filter.setStringLength(1);
filter.setMultipleKeys(alsoVisualize);
String message = Validation.validateForStringLevel(filter);
if (message == null) {

@ -10,6 +10,7 @@ import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.atomic.AtomicLong;
import data.Filter;
import data.MultipleHMKeys;
import org.apache.commons.csv.CSVFormat;
import org.apache.commons.csv.CSVPrinter;
import org.apache.commons.lang3.tuple.Pair;
@ -20,22 +21,22 @@ import data.Enums.WordLevelType;
@SuppressWarnings("unchecked")
public class Export {
public static void SetToJSON(Set<Pair<String, Map<String, Long>>> set) {
public static void SetToJSON(Set<Pair<String, Map<MultipleHMKeys, Long>>> set) {
JSONArray wrapper = new JSONArray();
for (Pair<String, Map<String, Long>> p : set) {
for (Pair<String, Map<MultipleHMKeys, Long>> p : set) {
JSONArray data_wrapper = new JSONArray();
JSONObject metric = new JSONObject();
String title = p.getLeft();
Map<String, Long> map = p.getRight();
Map<MultipleHMKeys, Long> map = p.getRight();
if (map.isEmpty())
continue;
long total = Util.mapSumFrequencies(map);
for (Map.Entry<String, Long> e : map.entrySet()) {
for (Map.Entry<MultipleHMKeys, Long> e : map.entrySet()) {
JSONObject data_entry = new JSONObject();
data_entry.put("word", e.getKey());
data_entry.put("frequency", e.getValue());
@ -56,8 +57,8 @@ public class Export {
}
}
public static String SetToCSV(Set<Pair<String, Map<String, Long>>> set, File resultsPath, LinkedHashMap<String, String> headerInfoBlock,
Map<String, Map<String, AtomicLong>> taxonomyResults) {
public static String SetToCSV(Set<Pair<String, Map<MultipleHMKeys, Long>>> set, File resultsPath, LinkedHashMap<String, String> headerInfoBlock,
Map<String, Map<MultipleHMKeys, AtomicLong>> taxonomyResults) {
//Delimiter used in CSV file
String NEW_LINE_SEPARATOR = "\n";
List<Object> FILE_HEADER_AL = new ArrayList<Object>();
@ -65,8 +66,8 @@ public class Export {
//Count frequencies
long num_frequencies = 0;
for (Pair<String, Map<String, Long>> p : set) {
Map<String, Long> map = p.getRight();
for (Pair<String, Map<MultipleHMKeys, Long>> p : set) {
Map<MultipleHMKeys, Long> map = p.getRight();
if (map.isEmpty())
continue;
num_frequencies = Util.mapSumFrequencies(map);
@ -88,21 +89,48 @@ public class Export {
if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("različnica")) {
headerInfoBlock.put("Skupna vsota vseh različnic:", String.valueOf(num_frequencies));
FILE_HEADER_AL.add("Različnica");
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse različnice");
} else if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("lema")) {
headerInfoBlock.put("Skupna vsota vseh lem:", String.valueOf(num_frequencies));
FILE_HEADER_AL.add("Lema");
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse leme");
} else if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("oblikoskladenjska oznaka")) {
headerInfoBlock.put("Skupna vsota vseh oblikoskladenjskih oznak:", String.valueOf(num_frequencies));
FILE_HEADER_AL.add("Oblikoskladenjska oznaka");
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse oblikoskladenjske oznake");
} else {
headerInfoBlock.put("Skupna vsota vseh različnic:", String.valueOf(num_frequencies));
FILE_HEADER_AL.add("Lema");
}
for (Map<MultipleHMKeys, AtomicLong> value : taxonomyResults.values()) {
for (MultipleHMKeys key : value.keySet()){
if(!key.getLemma().equals("")){
FILE_HEADER_AL.add("Lema");
}
if(!key.getWordType().equals("")){
FILE_HEADER_AL.add("Besedna vrsta");
}
if(!key.getMsd().equals("")){
FILE_HEADER_AL.add("Oblikoskladenjska oznaka");
}
break;
}
break;
}
if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("različnica")) {
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse različnice");
} else if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("lema")) {
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse leme");
} else if (headerInfoBlock.containsKey("Izračunaj za:") && headerInfoBlock.get("Izračunaj za:").equals("oblikoskladenjska oznaka")) {
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse oblikoskladenjske oznake");
} else {
FILE_HEADER_AL.add("Skupna absolutna pogostost");
FILE_HEADER_AL.add("Delež glede na vse leme");
}
@ -122,14 +150,14 @@ public class Export {
String fileName = "";
for (Pair<String, Map<String, Long>> p : set) {
for (Pair<String, Map<MultipleHMKeys, Long>> p : set) {
String title = p.getLeft();
fileName = title.replace(": ", "-");
fileName = fileName.replace(" ", "_").concat(".csv");
fileName = resultsPath.toString().concat(File.separator).concat(fileName);
Map<String, Long> map = p.getRight();
Map<MultipleHMKeys, Long> map = p.getRight();
if (map.isEmpty())
continue;
@ -155,9 +183,18 @@ public class Export {
//Create CSV file header
csvFilePrinter.printRecord(FILE_HEADER);
for (Map.Entry<String, Long> e : map.entrySet()) {
for (Map.Entry<MultipleHMKeys, Long> e : map.entrySet()) {
List dataEntry = new ArrayList<>();
dataEntry.add(e.getKey());
dataEntry.add(e.getKey().getKey());
if(!e.getKey().getLemma().equals("")){
dataEntry.add(e.getKey().getLemma());
}
if(!e.getKey().getWordType().equals("")){
dataEntry.add(e.getKey().getWordType());
}
if(!e.getKey().getMsd().equals("")){
dataEntry.add(e.getKey().getMsd());
}
dataEntry.add(e.getValue().toString());
dataEntry.add(formatNumberAsPercent((double) e.getValue() / num_frequencies));
dataEntry.add(String.format("%.2f", ((double) e.getValue() * 10000)/num_frequencies));

@ -9,6 +9,7 @@ import java.util.concurrent.TimeUnit;
import java.util.concurrent.atomic.AtomicLong;
import java.util.stream.Stream;
import data.MultipleHMKeys;
import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
@ -85,11 +86,11 @@ public class Util {
* Generic map converter -> since AtomicLongs aren't as comparable.
* Converts ConcurrentHashMap<K, AtomicLong> to HashMap<K, Long>
*/
public static <K, V> Map<String, Long> atomicInt2StringAndInt(Map<K, V> map) {
Map m = new HashMap<String, Long>();
public static <K, V> Map<MultipleHMKeys, Long> atomicInt2StringAndInt(Map<K, V> map) {
Map m = new HashMap<MultipleHMKeys, Long>();
for (Map.Entry<K, V> e : map.entrySet()) {
m.put(e.getKey().toString(), ((AtomicLong) e.getValue()).longValue());
m.put(e.getKey(), ((AtomicLong) e.getValue()).longValue());
}
return m;
@ -148,7 +149,7 @@ public class Util {
System.out.println();
}
static long mapSumFrequencies(Map<String, Long> map) {
static long mapSumFrequencies(Map<MultipleHMKeys, Long> map) {
long sum = 0;
for (long value : map.values()) {

Loading…
Cancel
Save