<br />

|---|
| [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/probability/blob/v0.25.0/tensorflow_probability/python/stats/sample_stats.py#L605-L641) |

Estimate variance using samples.

    tfp.stats.variance(
        x, sample_axis=0, keepdims=False, name=None
    )

Given `N` samples of scalar valued random variable `X`, variance may
be estimated as

    Var[X] := N^{-1} sum_{n=1}^N (X_n - Xbar) Conj{(X_n - Xbar)}
    Xbar := N^{-1} sum_{n=1}^N X_n

    x = tf.random.normal(shape=(100, 2, 3))

    # var[i, j] is the sample variance of the (i, j) batch member of x.
    var = tfp.stats.variance(x, sample_axis=0)

Notice we divide by `N` (the numpy default), which does not create `NaN`
when `N = 1`, but is slightly biased.

<br />

<br />

<br />

| ## Args ||
|---|---|
| `x` | A numeric `Tensor` holding samples. |
| `sample_axis` | Scalar or vector `Tensor` designating axis holding samples, or `None` (meaning all axis hold samples). Default value: `0` (leftmost dimension). |
| `keepdims` | Boolean. Whether to keep the sample axis as singletons. |
| `name` | Python `str` name prefixed to Ops created by this function. Default value: `None` (i.e., `'variance'`). |

<br />

<br />

<br />

<br />

| ## Returns ||
|---|---|
| `var` | A `Tensor` of same `dtype` as the `x`, and rank equal to `rank(x) - len(sample_axis)` |

<br />