Remove

Name	Mandatory	Description	Default	Type
⬅️ Input		Any input is ignored.		None
Output ➡️		Output is the filtered sequence.		[Any]
From	No	The name of the sequence variable to edit in place.	none	Var([Any])
Join	No	Other columns to join sort/filter using the input (they must be of the same length).	none	Var([Any])[Var([Any])]
Predicate	No	The shards to use as predicate, if true the item will be popped from the sequence.	none	Shard[Shard]
Unordered	No	Turn on to remove items very quickly but will not preserve the sequence items order.	false	Bool

Removes all elements from a sequence that match the given condition. Can also take these matched indices and remove corresponding elements from a joined sequence.

Details

Remove removes all the elements of the Name parameter sequence that matches the condition laid out in the Predicate parameter.

It can also take these condition-matched indices (from the From sequence) and remove the corresponding elements from a joined sequence (passed via the Join parameter). Remember, Remove doesn't apply the Predicate conditions to the joined sequence, but removes corresponding elements from it based on Predicate matched indices of the main sequence. For this to work both the sequences must have the same length.

[!NOTE] Think of this as the Shards equivalent of a relational database inner join. The main sequence and the joined sequence can be thought of as columns from two different tables inner joined over indices equality. So that the changes in elements of one sequence (rows in the first table) can be propagated to the corresponding elements of the joined sequence (corresponding rows in the joined table).

In this case the operation is deletion of selected elements (selected rows) from one sequence (table) leading to deletion of corresponding elements (connected rows) of the joined sequence (joined table).

The Predicate parameter can take any conditional/logical expression or combination of shards that will result in assertion that can be tested on the sequence elements.

The Unordered parameter can be set to true if you need to make the removal process faster, but then the order of the remaining elements in the resulting sequence elements may not be preserved. By default, this order is preserved.

Remove works only on sequences.

Any input to this shard is ignored and its output is the main filtered sequence.

Examples

Code

; ; remove predicate-satisfying elements from From: sequence
[10 20 30] >= seq1
Remove(
  From: seq1
  Predicate: IsMore(20)
  Unordered: false
)
Log("output") ; ; => output: [10, 20]
seq1 | Log("seq1") ; ; index-2 element matched, removed => seq1: [10, 20]

; ; remove corresponding index elements from Join: sequence
[10 20 30] >= seq2
[100 200 300] >= seqJ
Remove(seq2 Predicate: Is(20) Join: seqJ)
seq2 | Log("seq2") ; ; index-1 element matched, removed from seq2 => seq2: [10, 30]
seqJ | Log("seqJ") ; ; matched index-1 element removed from seqJ  => seqJ: [100, 300]

; ; remove with Unordered: true (faster but sequence items order may not be preserved)
[[20 30] [30 40] [40 50]] >= seq3
Remove(seq3 Predicate: {Take(0) | IsLess(30)} Unordered: true)
seq3 | Log("seq3")

 

---