---
file_format: mystnb
kernelspec:
  name: python3
---

(mutual_information)=

# Mutual Information

Mutual Information (MI) measures the amount of information obtained about one random
variable through observing another random variable.

```python
import delaynet as dn

# Calculate mutual information
result = dn.connectivity(
    ts1, ts2, metric="mutual_information", approach="kernel", lag_steps=5
)
```

or more involved with ordinal mutual information

```python
import delaynet as dn

# Calculate transfer entropy
result = dn.connectivity(
    ts1,
    ts2,
    metric="mutual_information",
    approach='ordinal',
    lag_steps=5,
    embedding_dim=3
)
```

Parameters:

- `approach`: The approach to use for estimating mutual information.
  For available approaches, find the string in {func}`infomeasure.mutual_information`.
- `lag_steps`: Time lags to consider. An integer will consider lags [1, ..., lag_steps].
  Passing a list will consider the specified values as lags.
- `hypothesis_type`: Type of hypothesis test to use ('permutation_test' or 'bootstrap').
  Default is 'permutation_test'.
- `n_tests`: Number of iterations or resamples to perform within the hypothesis test.
  Default is 20.
- Additional keyword arguments are passed directly to the mutual information estimator.

```{eval-rst}
.. automethod:: delaynet.connectivities.mutual_information
    :noindex:
```