Merge pull request #617 from CartoDB/497-null-category

Be able to not compute NULL categories and null values
This commit is contained in:
Daniel
2017-01-20 11:26:12 +01:00
committed by GitHub
3 changed files with 130 additions and 34 deletions
+33 -10
View File
@@ -19,25 +19,37 @@ var rankedCategoriesQueryTpl = dot.template([
' SELECT {{=it._column}} AS category, {{=it._aggregation}} AS value,',
' row_number() OVER (ORDER BY {{=it._aggregation}} desc) as rank',
' FROM ({{=it._query}}) _cdb_aggregation_all',
' {{?it._aggregationColumn!==null}}WHERE {{=it._aggregationColumn}} IS NOT NULL{{?}}',
' GROUP BY {{=it._column}}',
' ORDER BY 2 DESC',
')'
].join('\n'));
var categoriesSummaryQueryTpl = dot.template([
'categories_summary AS(',
' SELECT count(1) categories_count, max(value) max_val, min(value) min_val',
var categoriesSummaryMinMaxQueryTpl = dot.template([
'categories_summary_min_max AS(',
' SELECT max(value) max_val, min(value) min_val',
' FROM categories',
')'
].join('\n'));
var categoriesSummaryCountQueryTpl = dot.template([
'categories_summary_count AS(',
' SELECT count(1) AS categories_count',
' FROM (',
' SELECT {{=it._column}} AS category',
' FROM ({{=it._query}}) _cdb_categories',
' GROUP BY {{=it._column}}',
' ) _cdb_categories_count',
')'
].join('\n'));
var rankedAggregationQueryTpl = dot.template([
'SELECT CAST(category AS text), value, false as agg, nulls_count, min_val, max_val, count, categories_count',
' FROM categories, summary, categories_summary',
' FROM categories, summary, categories_summary_min_max, categories_summary_count',
' WHERE rank < {{=it._limit}}',
'UNION ALL',
'SELECT \'Other\' category, sum(value), true as agg, nulls_count, min_val, max_val, count, categories_count',
' FROM categories, summary, categories_summary',
' FROM categories, summary, categories_summary_min_max, categories_summary_count',
' WHERE rank >= {{=it._limit}}',
'GROUP BY nulls_count, min_val, max_val, count, categories_count'
].join('\n'));
@@ -45,7 +57,7 @@ var rankedAggregationQueryTpl = dot.template([
var aggregationQueryTpl = dot.template([
'SELECT CAST({{=it._column}} AS text) AS category, {{=it._aggregation}} AS value, false as agg,',
' nulls_count, min_val, max_val, count, categories_count',
'FROM ({{=it._query}}) _cdb_aggregation_all, summary, categories_summary',
'FROM ({{=it._query}}) _cdb_aggregation_all, summary, categories_summary_min_max, categories_summary_count',
'GROUP BY category, nulls_count, min_val, max_val, count, categories_count',
'ORDER BY value DESC'
].join('\n'));
@@ -114,6 +126,7 @@ Aggregation.prototype.sql = function(psql, override, callback) {
var _query = this.query;
var aggregationSql;
if (!!override.ownFilter) {
aggregationSql = [
"WITH",
@@ -125,9 +138,14 @@ Aggregation.prototype.sql = function(psql, override, callback) {
rankedCategoriesQueryTpl({
_query: _query,
_column: this.column,
_aggregation: this.getAggregationSql()
_aggregation: this.getAggregationSql(),
_aggregationColumn: this.aggregation !== 'count' ? this.aggregationColumn : null
}),
categoriesSummaryQueryTpl({
categoriesSummaryMinMaxQueryTpl({
_query: _query,
_column: this.column
}),
categoriesSummaryCountQueryTpl({
_query: _query,
_column: this.column
})
@@ -150,9 +168,14 @@ Aggregation.prototype.sql = function(psql, override, callback) {
rankedCategoriesQueryTpl({
_query: _query,
_column: this.column,
_aggregation: this.getAggregationSql()
_aggregation: this.getAggregationSql(),
_aggregationColumn: this.aggregation !== 'count' ? this.aggregationColumn : null
}),
categoriesSummaryQueryTpl({
categoriesSummaryMinMaxQueryTpl({
_query: _query,
_column: this.column
}),
categoriesSummaryCountQueryTpl({
_query: _query,
_column: this.column
})
@@ -18,25 +18,37 @@ var rankedCategoriesQueryTpl = dot.template([
' SELECT {{=it._column}} AS category, {{=it._aggregation}} AS value,',
' row_number() OVER (ORDER BY {{=it._aggregation}} desc) as rank',
' FROM ({{=it._query}}) _cdb_aggregation_all',
' {{?it._aggregationColumn!==null}}WHERE {{=it._aggregationColumn}} IS NOT NULL{{?}}',
' GROUP BY {{=it._column}}',
' ORDER BY 2 DESC',
')'
].join('\n'));
var categoriesSummaryQueryTpl = dot.template([
'categories_summary AS(',
' SELECT count(1) categories_count, max(value) max_val, min(value) min_val',
var categoriesSummaryMinMaxQueryTpl = dot.template([
'categories_summary_min_max AS(',
' SELECT max(value) max_val, min(value) min_val',
' FROM categories',
')'
].join('\n'));
var categoriesSummaryCountQueryTpl = dot.template([
'categories_summary_count AS(',
' SELECT count(1) AS categories_count',
' FROM (',
' SELECT {{=it._column}} AS category',
' FROM ({{=it._query}}) _cdb_categories',
' GROUP BY {{=it._column}}',
' ) _cdb_categories_count',
')'
].join('\n'));
var rankedAggregationQueryTpl = dot.template([
'SELECT CAST(category AS text), value, false as agg, nulls_count, min_val, max_val, count, categories_count',
' FROM categories, summary, categories_summary',
' FROM categories, summary, categories_summary_min_max, categories_summary_count',
' WHERE rank < {{=it._limit}}',
'UNION ALL',
'SELECT \'Other\' category, sum(value), true as agg, nulls_count, min_val, max_val, count, categories_count',
' FROM categories, summary, categories_summary',
' FROM categories, summary, categories_summary_min_max, categories_summary_count',
' WHERE rank >= {{=it._limit}}',
'GROUP BY nulls_count, min_val, max_val, count, categories_count'
].join('\n'));
@@ -44,7 +56,7 @@ var rankedAggregationQueryTpl = dot.template([
var aggregationQueryTpl = dot.template([
'SELECT CAST({{=it._column}} AS text) AS category, {{=it._aggregation}} AS value, false as agg,',
' nulls_count, min_val, max_val, count, categories_count',
'FROM ({{=it._query}}) _cdb_aggregation_all, summary, categories_summary',
'FROM ({{=it._query}}) _cdb_aggregation_all, summary, categories_summary_min_max, categories_summary_count',
'GROUP BY category, nulls_count, min_val, max_val, count, categories_count',
'ORDER BY value DESC'
].join('\n'));
@@ -85,9 +97,14 @@ Aggregation.prototype.sql = function(psql, override, callback) {
rankedCategoriesQueryTpl({
_query: _query,
_column: this.column,
_aggregation: this.getAggregationSql()
_aggregation: this.getAggregationSql(),
_aggregationColumn: this.aggregation !== 'count' ? this.aggregationColumn : null
}),
categoriesSummaryQueryTpl({
categoriesSummaryMinMaxQueryTpl({
_query: _query,
_column: this.column
}),
categoriesSummaryCountQueryTpl({
_query: _query,
_column: this.column
})
@@ -110,9 +127,14 @@ Aggregation.prototype.sql = function(psql, override, callback) {
rankedCategoriesQueryTpl({
_query: _query,
_column: this.column,
_aggregation: this.getAggregationSql()
_aggregation: this.getAggregationSql(),
_aggregationColumn: this.aggregation !== 'count' ? this.aggregationColumn : null
}),
categoriesSummaryQueryTpl({
categoriesSummaryMinMaxQueryTpl({
_query: _query,
_column: this.column
}),
categoriesSummaryCountQueryTpl({
_query: _query,
_column: this.column
})
+65 -14
View File
@@ -3,7 +3,7 @@ require('../../support/test_helper');
var assert = require('../../support/assert');
var TestClient = require('../../support/test-client');
describe('aggregations', function() {
describe('aggregations happy cases', function() {
afterEach(function(done) {
if (this.testClient) {
@@ -13,30 +13,36 @@ describe('aggregations', function() {
}
});
function aggregationOperationMapConfig(operation) {
return {
function aggregationOperationMapConfig(operation, query, column, aggregationColumn) {
column = column || 'adm0name';
aggregationColumn = aggregationColumn || 'pop_max';
query = query || 'select * from populated_places_simple_reduced';
var mapConfig = {
version: '1.5.0',
layers: [
{
type: 'mapnik',
options: {
sql: 'select * from populated_places_simple_reduced',
sql: query,
cartocss: '#layer0 { marker-fill: red; marker-width: 10; }',
cartocss_version: '2.0.1',
widgets: {
adm0name: {
type: 'aggregation',
options: {
column: 'adm0name',
aggregation: operation,
aggregationColumn: 'pop_max'
}
}
}
widgets: {}
}
}
]
};
mapConfig.layers[0].options.widgets[column] = {
type: 'aggregation',
options: {
column: column,
aggregation: operation,
aggregationColumn: aggregationColumn
}
};
return mapConfig;
}
var operations = ['count', 'sum', 'avg', 'max', 'min'];
@@ -56,4 +62,49 @@ describe('aggregations', function() {
});
});
});
var query = [
'select 1 as val, \'a\' as cat, ST_Transform(ST_SetSRID(ST_MakePoint(0,0),4326),3857) as the_geom_webmercator',
'select null, \'b\', ST_Transform(ST_SetSRID(ST_MakePoint(0,1),4326),3857)',
'select null, \'b\', ST_Transform(ST_SetSRID(ST_MakePoint(1,0),4326),3857)',
'select null, null, ST_Transform(ST_SetSRID(ST_MakePoint(1,1),4326),3857)'
].join(' UNION ALL ');
operations.forEach(function (operation) {
var not = operation === 'count' ? ' not ' : ' ';
var description = 'should' +
not +
'handle NULL values in category and aggregation columns using "' +
operation +
'" as aggregation operation';
it(description, function (done) {
this.testClient = new TestClient(aggregationOperationMapConfig(operation, query, 'cat', 'val'));
this.testClient.getDataview('cat', { own_filter: 0 }, function (err, aggregation) {
assert.ifError(err);
assert.ok(aggregation);
assert.equal(aggregation.type, 'aggregation');
assert.ok(aggregation.categories);
assert.equal(aggregation.categoriesCount, 3);
assert.equal(aggregation.count, 4);
assert.equal(aggregation.nulls, 1);
var hasNullCategory = false;
aggregation.categories.forEach(function (category) {
if (category.category === null) {
hasNullCategory = true;
}
});
if (operation === 'count') {
assert.ok(hasNullCategory, 'aggregation has not a category NULL');
} else {
assert.ok(!hasNullCategory, 'aggregation has category NULL');
}
done();
});
});
});
});