# tf_agents.policies.utils.masked_argmax

<br />

|---|
| [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/agents/blob/v0.19.0/tf_agents/policies/utils.py#L225-L254) |

Computes the argmax where the allowed elements are given by a mask.

    tf_agents.policies.utils.masked_argmax(
        input_tensor: https://www.tensorflow.org/agents/api_docs/python/tf_agents/typing/types/Tensor,
        mask: https://www.tensorflow.org/agents/api_docs/python/tf_agents/typing/types/Tensor,
        output_type: tf.DType = tf.int32
    ) -> https://www.tensorflow.org/agents/api_docs/python/tf_agents/typing/types/Tensor

If a row of `mask` contains all zeros, then this method will return -1 for the
corresponding row of `input_tensor`.

<br />

<br />

<br />

| ## Args ||
|---|---|
| `input_tensor` | Rank-2 Tensor of floats. |
| `mask` | 0-1 valued Tensor of the same shape as input. |
| `output_type` | Integer type of the output. |

<br />

<br />

<br />

<br />

| ## Returns ||
|---|---|
| A Tensor of rank 1 and type `output_type`, with the masked argmax of every row of `input_tensor`. ||

<br />