# `Latu.ML.Feature.StringIndexerModel`
[🔗](https://github.com/zero-one-group/latu_ml/blob/v0.4.2/lib/latu/ml/feature.ex#L0)

Attributes of a fitted `StringIndexerModel`.

Every accessor here is one name on the server's allowlist for
`org.apache.spark.ml.feature.StringIndexerModel` — 2 of them — so tab completion is that
allowlist. The server refuses anything else with `CONNECT_ML.ATTRIBUTE_NOT_ALLOWED`; these
refuse a model of the wrong class before it gets that far, and name the module that would
have taken it.

Allowed by the server but not here: `toString` is `Identifiable`'s, and tells you less than
the model's own uid. `Latu.ML.attribute/2` and `Latu.ML.attribute/3` will send any of these
names; it is what comes back that has nowhere to go.

This model does not have to be fitted: `from_labels/3`, `from_arrays_of_labels/3` builds one
on the server out of what you pass, through its own helper method. That is a cache entry
like a fit's, and yours to give back.

The allowlist is inherited rather than per class: this one is the union of
`org.apache.spark.ml.feature.StringIndexerModel`, `org.apache.spark.ml.util.Identifiable`.

# `class`

```elixir
@spec class() :: String.t()
```

The JVM class these accessors belong to.

# `from_arrays_of_labels`

```elixir
@spec from_arrays_of_labels(Latu.Session.t(), [[String.t() | atom()]], keyword()) ::
  {:ok, Latu.ML.Model.t()} | {:error, Latu.Error.t()}
```

Construct the model directly from an array of array of label strings, requires an active
SparkContext. The server builds it and registers it, so what comes back is a cache reference
like a fit's and is yours to `Latu.ML.delete/1`. The uid is generated here and sent with the
call. `opts` are params set on the model afterwards, checked against its own table. *Built
by* `stringIndexerModelFromLabelsArray` on the server's helper object, from
`StringIndexerModel.from_arrays_of_labels`.

# `from_labels`

```elixir
@spec from_labels(Latu.Session.t(), [String.t() | atom()], keyword()) ::
  {:ok, Latu.ML.Model.t()} | {:error, Latu.Error.t()}
```

Construct the model directly from an array of label strings, requires an active
SparkContext. The server builds it and registers it, so what comes back is a cache reference
like a fit's and is yours to `Latu.ML.delete/1`. The uid is generated here and sent with the
call. `opts` are params set on the model afterwards, checked against its own table. *Built
by* `stringIndexerModelFromLabels` on the server's helper object, from
`StringIndexerModel.from_labels`.

# `labels`

```elixir
@spec labels(Latu.ML.Model.t()) :: {:ok, term()} | {:error, Latu.Error.t()}
```

Ordered list of labels, corresponding to indices to be assigned.

An action: it reaches the server. A `Vector` or `Matrix` comes back as an `Nx.Tensor`, or a
`Latu.ML.SparseVector` where densifying would be this package's decision rather than yours.

# `labels_array`

```elixir
@spec labels_array(Latu.ML.Model.t()) :: {:ok, term()} | {:error, Latu.Error.t()}
```

Array of ordered list of labels, corresponding to indices to be assigned for each input
column.

An action: it reaches the server. A `Vector` or `Matrix` comes back as an `Nx.Tensor`, or a
`Latu.ML.SparseVector` where densifying would be this package's decision rather than yours.

---

*Consult [api-reference.md](api-reference.md) for complete listing*