# `:` — Unkey / Keyed

Function. Ranks: `∞` monadic, `1 ∞` dyadic

`:Y` removes axis keys. `K:Y` labels leading axes. `K:⍠A Y` labels the
axes `A`. A bracket list with any `key:value` item is one keyed vector.
The list’s other items are entries without keys. See [axis
keys](../keyed.ipynb).

``` bpl
:"aa" "bb":1 2    ⍝ 1 2
"bb"⊃"aa" "bb":1 2 ⍝ 2
["aa":1 "bb":2]≡"aa" "bb":1 2 ⍝ $t
```

In `K`, a position’s own number leaves that position without a key. Axis
names are the keys of the shape. `N:⍴M ⍴ M` renames the axes. See
[shape](rho.qmd).

``` bpl
"aa" 1:10 20        ⍝ ["aa":10 20]
M←["city":2 "month":3]⍴⍳6
⍴M⍴⍨:⍴M            ⍝ [2 3]ₓ
```

A keyed `K` supplies axis names and position keys together.

``` bpl
axes←["city":"NY" "LA";"month":"Jan" "Feb" "Mar"]
M←axes:[1 2 3⋄4 5 6]
"LA" "Feb"⌷M       ⍝ 5
```

## Inverse

`K:⁻¹Y` removes the keys `K` from `Y`, which must have exactly those
keys.

``` bpl
"a" "b":⁻¹["a":1;"b":2]   ⍝ 1 2
```