# `SparkEx.DataFrame.NA`
[🔗](https://github.com/lukaszsamson/spark_ex/blob/v0.1.1/lib/spark_ex/data_frame/na.ex#L1)

Null-value handling sub-API for DataFrames.

Provides `fill/2`, `drop/1`, and `replace/3` operations that return
new lazy DataFrames with the corresponding NA plan tuples.

Accessed via `SparkEx.DataFrame.fillna/2`, `SparkEx.DataFrame.dropna/1`,
`SparkEx.DataFrame.replace/3`, or directly.

# `drop`

```elixir
@spec drop(
  SparkEx.DataFrame.t(),
  keyword()
) :: SparkEx.DataFrame.t()
```

Drops rows containing null values.

## Options

- `:how` — `:any` (default) drops rows with any null; `:all` drops rows where all values are null.
- `:thresh` — minimum number of non-null values required to keep a row.
  Overrides `:how` when provided.
- `:subset` — list of column names to consider.

## Examples

    DataFrame.NA.drop(df)
    DataFrame.NA.drop(df, how: :all)
    DataFrame.NA.drop(df, thresh: 2, subset: ["age", "name"])

# `fill`

```elixir
@spec fill(SparkEx.DataFrame.t(), term(), keyword()) :: SparkEx.DataFrame.t()
```

Fills null values with the given replacement.

## Parameters

- `value` — scalar (int, float, string, bool) to fill all null values,
  or a map `%{"column_name" => replacement_value}` for column-specific fills.
- `opts` — keyword options:
  - `:subset` — list of column names to restrict the fill to.

## Examples

    DataFrame.NA.fill(df, 0)
    DataFrame.NA.fill(df, %{"age" => 0, "name" => "unknown"})
    DataFrame.NA.fill(df, 0, subset: ["age", "salary"])

# `replace`

```elixir
@spec replace(SparkEx.DataFrame.t(), term(), term(), keyword()) ::
  SparkEx.DataFrame.t()
```

Replaces values in the DataFrame.

## Forms

- `replace(df, %{old => new, ...})` — replacement map
- `replace(df, old_value, new_value)` — single replacement
- `replace(df, [old1, old2], [new1, new2])` — parallel lists

## Options

- `:subset` — list of column names to restrict replacements to.

## Examples

    DataFrame.NA.replace(df, %{0 => 100, -1 => 0})
    DataFrame.NA.replace(df, "N/A", nil)
    DataFrame.NA.replace(df, [1, 2], [10, 20], subset: ["score"])

---

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