python.qmd

---
title: "Python"
---

The [`rembus`](https://pypi.org/project/rembus/) Python package provide both a synchronous and an
asynchronous interface.

```python
# async API
import rembus
```

```python
# sync API
import rembus.sync as rembus
```

## Getting Started

```bash
pip install rembus
```

Then create a component object to interact with the others components:

```python
# sync API
rb = rembus.component()

# async API
rb = await rembus.component()
```

The `rb` object provides methods for exposing functions implementation, for subscribing to topics, for message publishing and for service requests:

- rpc
- publish
- expose
- subscribe

## Request a service (RPC)

This example is a demo of requesting the service [`stats`](julia.qmd#implements-a-service-rpc) implemented in Julia.

- A python pandas dataframe is created;
- The pandas dataframe is used as an argument of the rpc service `stats`;
- On the RPC server side the `stats` method receive a Julia DataFrame;
- The return value of `stats` is a Julia Dictionary;
- On the client RPC side the response is a python dictionary.

```python
import rembus.sync as rembus
from random import random
import pandas as pd

df = pd.DataFrame({
    "name": [f"kpi_{i}" for i in range(1,6)],
    "ts": range(1,6),
    "value": [random() for i in range(1,6)]
})

rb = rembus.component()
summary = rb.rpc("stats", df)

```