conf.rb

# frozen_string_literal: true
require 'deep-freeze'

# DIMENSIONS & GROUPS

# A dimension will get aggregated based on unique values. In final output
# `key` will be value of the dimension. By default number of matched documents
# will be the `value`.
#
# You can specify special aggregations using Elastic Aggregation syntax.
# https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations.html
#
# It is likely that dimensions will be based on term
# https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations-bucket-terms-aggregation.html

# These will get converted to term aggregations like
# { terms: { field: 'year', size: 10000 } }
#
# Each Dimension must be a field as per index definition.

# CUSTOM AGGREGATIONS (similar to Crossfilter custom Groups)

# The custom aggregations will return an Object
#       {
#         "key": "Q1",
#         "value": {
#           "doc_count": 1668,
#           "volume": 18819625217.0
#         }
#       }
# To only get the value like the following, use `only_me` as the aggregation name
#       {
#         "key": "Q1",
#         "value": 18819625217.0
#       }
# Equivalent to the following from stock.js
# quarter.group().reduceSum(d => d.volume)
quarter_chart_agg = { only_me: { sum: { field: 'volume' } } }.deep_freeze

# moveMonths.group().reduceSum(d => Math.abs(d.close - d.open))
index_move_by_month = { only_me: { sum: { script: { source: 'Math.abs( doc.close.value - doc.open.value )' } } } }.deep_freeze

# moveMonths.group().reduceSum(d => d.volume / 500000);
# Since volume is integer as per index definition, it is important to use 500000.0 to avoid integer division
volume_by_month_group_agg = { only_me: { sum: { script: { source: 'doc.volume.value / 500000.0' } } } }.deep_freeze

# ++p.days;
# p.total += (v.open + v.close) / 2;
# p.avg = Math.round(p.total / p.days);
#
# Elastic has concept of pipeline aggregations
# (https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations-pipeline.html)
# These help in computing derived values based on other aggregations.
# For example, in this case after computing average, it is rounded
index_avg_by_month_agg = {
  avg_without_rounding: {
    avg: {
      script: {
        source: '( doc.open.value + doc.close.value ) / 2',
      },
    },
  },
  avg: {
    bucket_script: {
      buckets_path: {
        avg_without_rounding: 'avg_without_rounding',
      },
      script: 'Math.round(params.avg_without_rounding)',
    },
  },
}.deep_freeze

# creating aggregation for these (from stock.js example)
# p.absGain += v.close - v.open;
# p.fluctuation += Math.abs(v.close - v.open);
# p.sumIndex += (v.open + v.close) / 2;
# p.avgIndex = p.sumIndex / p.count;
# p.percentageGain = p.avgIndex ? (p.absGain / p.avgIndex) * 100 : 0;
# p.fluctuationPercentage = p.avgIndex ? (p.fluctuation / p.avgIndex) * 100 : 0;
yearly_bubble_chart_agg = {
  avgIndex: {
    avg: {
      script: {
        source: '( doc.open.value + doc.close.value ) / 2',
      },
    },
  },
  absGain: {
    sum: {
      script: {
        source: '( doc.close.value - doc.open.value )',
      },
    },
  },
  fluctuation: {
    sum: {
      script: {
        source: 'Math.abs( doc.close.value - doc.open.value )',
      },
    },
  },
  percentageGain: {
    bucket_script: {
      buckets_path: {
        avgIndex: 'avgIndex',
        absGain: 'absGain',
      },
      script: 'params.absGain / params.avgIndex * 100',
    },
  },
  fluctuationPercentage: {
    bucket_script: {
      buckets_path: {
        avgIndex: 'avgIndex',
        fluctuation: 'fluctuation',
      },
      script: 'params.fluctuation / params.avgIndex * 100',
    },
  },
}.deep_freeze

# Only the CONF is used outside this file
CONF = {
  # whether to include selected/total row-counts
  counts: true,
  index: 'stocks',
  # dim_id as key and then group_id as key and aggregation as value
  # use nil as aggregation if we need the default group (the matched count)
  'dims_and_groups' => {
    'year' => {
      'yearly_performance' => yearly_bubble_chart_agg,
    },
    'gain_or_loss' => {
      'gain_or_loss' => nil,
    },
    'quarter' => {
      'quarter' => quarter_chart_agg,
    },
    'day_of_week' => {
      'day_of_week' => nil,
    },
    'fluctuation' => {
      'fluctuation' => nil,
    },
    'month' => {
      'index_avg_by_month' => index_avg_by_month_agg,
      'index_move_by_month' => index_move_by_month,
      'volume_by_month' => volume_by_month_group_agg,
    },
  },
}.deep_freeze # make this Object immutable