gupalo/arrayutils

Array utils

1.10.0 2024-01-26 18:33 UTC

Requires

• php: >=8.0
• symfony/property-access: ^5.0|^6.0|^7.0

Requires (Dev)

• phpunit/phpunit: ^10.0

MIT

README

Install

composer require gupalo/arrayutils

Features

All code is covered by tests. Look at tests to understand better how it's working.

All methods are static.

Items can be iterable (array, \Generator, ...) of one of:

• array
• object with public properties or getters (anything Symfony PropertyAccessor can get)

ArrayAggregator

• max(items, field): max field value among items
• sum(items, field): sum of field values of items
• ratio(items, field, field2: "sum by field"/"sum by field2"
• maxRatio(items, field, field2): max among "field value"/"field value2"

ArrayComparer

Comparing is smart - if compared values are float, they are rounded with precision = 8.

• compareOne(a1, a2, keys): check that each (string)a1[key] === (string)a2[key] for each of keys. returns "patch" how to go from a1 to a2. null value means that we need to unset this key.
• compare(a1, a2, keys) a1 and a2 are expected to be array of arrays. returns patch how what's changed in a2 compared to a1 - array with keys:
  • created: which items should be created and their values
  • updated: which items should be updated and updated fields
  • removed: which items (ids) should be deleted
• compareFlat(a1, a2) similar to compare but a1 and a2 are considered scalars and are compared fully, not by fields

ArrayFactory

• createKeys(keys, value): create array with these keys and value
• createDictionary(a, keyColumn, valueColumn): similar to array_column(a, valueColumn, keyColumn) but doesn't require values present in all items (?? null)

ArrayKeysHelper

• index($a, array|string $keys = 'id'): create indexed array where index may contain several fields
• indexAndGroup($a, array|string $keys = 'id'): create indexed array but allow several items with same index
• fill($a, $keys, $defaultValue = null): ensure that all keys exist in array and all others are missing
• filter($a, $keys, $createMissingKeys = true): get these keys from array
• unset($a, $keys): unset these keys from each array item

ArrayRandom

• pick(array $items): select random value from array
• pickMultiple(array $items, int $count = 1, bool $preserveKeys = false, mixed $default = []): select multiple ($count) random values from array

ArrayTable

• arrayToKeyValues(array $data): covert table (like TSV; first item is header with keys) to array with named keys

ArrayUniquer

• values(array $a): uniq -> values
• notNullValues(array $a): filter not null -> uniq -> values
• columnValues(array $a, string $column): column -> uniq -> values
• mergeValues(...$a): merge -> uniq -> values
• mergeNotNullValues(...$a): merge -> filter not null -> uniq -> values

FloatArrayFactory

• createPreserveKeys(array $a): ensure that all values are float
• create(array $a): ensure that all values are float -> values

IntArrayFactory

• createPreserveKeys(array $a): ensure that all values are int
• create(array $a): ensure that all values are int -> values

Balancer

Problem to solve: you have some items and need to split them to some buckets. You may configure different strategies how to distribute your items between the buckets. Buckets should implement BucketInterface, balancer should implement BalancerInterface

SemiRandomBalancer

It's not fully random, you set seed so multiple runs with the same seed and same items will lead to the same result.

FairBalancer

Try to make same weighted count of items in same bucket.

Example: A/B test where you want to send 90% of users to "A" page and 10% to "B" page. You could use simple random but with low amount of users it's quite probable that you have 20 users and all will see page "A" or 5 users that go to "B". Then you'll have hard time explaining probability theory. With this balancer you may be sure that 1 of the first 10 users will go to "B" and most probably 2 of 20 users will land there.