markharwood

curl -XGET "http://localhost:9200/meetuprsvps/_xpack/_graph/_explore" -d'
{
  "query": {
    "term": {
      "_all": "elasticsearch"
    }
  },
  "vertices": [
    {
      "field": "member_id",

markharwood

        TransportClient esClient = new TransportClient(ImmutableSettings.settingsBuilder().put("cluster.name", clusterName).build());

        // ======= Start hack======
        Field f = esClient.getClass().getDeclaredField("injector");
        f.setAccessible(true);
        Injector injector = (Injector) f.get(esClient);
        injector.createChildInjector(new SignificantTermsHeuristicModule());
        // ======= End hack======

markharwood

import gzip
import csv
import time
import datetime
import json

 # This script reads a data file (car test results) and uses it to create a CSV file of query clauses for benchmarking.
 # It creates mostly well-behaved queries and randomly throws in "problem" queries of various forms
 # Each CSV row in the output is a set of clauses (query/agg/filter) and some metadata for reporting purposes.

markharwood

import gzip
import csv
from elasticsearch import helpers
from elasticsearch.client import Elasticsearch
import time

csv.register_dialect('piper', delimiter='|', quoting=csv.QUOTE_NONE)

# See http://postcodepal.com/dbgen/postcode_areas_true-centroids.zip 
pf=open('/Users/Mark/Documents/work/irdata/MOT/postcode_areas.csv')

markharwood

//import org.apache.lucene.codecs.bloom.FuzzySet;

// Extract the doc source to a field
doc = ctx._source;

// Convert basic array into map for ease of manipulation
tagMap = doc.tags.collectEntries{[it.tag, it]};


// Patch the new tags into the data structure, adding one to a usercount

markharwood

{
  "title": "eslogs",
  "services": {
    "query": {
      "list": {
        "0": {
          "query": "",
          "alias": "",
          "color": "#7EB26D",
          "id": 0,

markharwood

//================================
// Here a script for gathering the precision/recall stats for a query (see http://en.wikipedia.org/wiki/Precision_and_recall)
// A candidate classifier query's effectiveness is determined by counting hits on pre-classified content
// If we compute the F-measure we can potentially use it as the fitness function for a genetic algo that mutates our query
// (introducing phrases, minShouldMatch clauses etc) to move us towards our target goal of balancing precision/recall in our classifier.
//=================================

// Our candidate query for classifying documents in a category
var candidateQuery={ "terms": {"body":    ["vs", "shr", "cts", "net", "revs", "note", "loss", "mths", "shrs", "avg", "profit"]}};
// Our filter criteria for identifying documents in our target category

markharwood

#!/bin/sh
es_host="http://localhost:9200"
curl -XDELETE "$es_host/contests"
curl -XPUT "$es_host/contests" -d '
{
   "settings": {
      "index.number_of_replicas": 0,
      "index.number_of_shards": 1,
      "index.refresh_interval": -1
   },

markharwood

//This function was used in my test rig to convert elasticsearch results to a KML structure which is
// later fed to an iFrame wrapping the GoogleEarth plugin

var data=[];
var buckets=inData.aggregations.map.buckets;

function addCommas(nStr)
{
    nStr += '';

markharwood

    //Groups a related set of terms, typically from the results of some auto-expansion,
    // and provides the average DocFreq of the set in order to avoid Lucene's IDF ranking
    // favouring the rarest interpretation, which is often a poor choice for auto-expanded
    // terms e.g. the terms produced by a fuzzy query or trying alternative fields
    class CommonIDFContext {
        int commonDf = -1;
        Map<Term, Integer> balancedDfs;
        List<Term> commonTerms = new ArrayList<Term>();

        public void add(Term unbalancedQueryTerm) {