Python ​

The Python widget integrates client-side Python to enable advanced analytics and visualizations in Dashboards. The Python widget is based on Pyodide library to run Python code in the browser. This means that you don't depend on any server and also increases privacy. The main disadvantage is scalability and the availability of packages. That said, if you need more we have Python, .Net and NodeJs backend services in our enterprise offering. Contact us for more info.

Python packages ​

Pyodide comes with a large amount of packages pre-installed. You can check the list of available packages here. If you need a package that is not available, let us know.

The most important packages included are: numpy, pandas, scikit-learn, matplotlib, scipy, networkx and igraph.

Qwiery package ​

The Python widget integrates with Qwiery via the Q package. This package is a thin wrapper around the Qwiery API and allows you to interact with the Qwiery backend. The Q package is automatically available in the Python widget, but you can also import it explicitly via

import Q

Other packages can be imported as usual. For example, if you want to use networkx to create an Erdos-Renyi graph, you can do:

import networkx as nx
g=nx.erdos_renyi_graph(12,0.5)
gs ={
    "nodes": list(g.nodes()),
    "edges":list(g.edges())
}
Q.graph(gs)

Several things should be noted here:

• the Q.graph function is used to send the graph to the dashboard widget. This will result in a graph being rendered.
• the nx.Graph graph object is converted to a dictionary with nodes and edges. This is necessary because the graph object is not serializable and cannot be sent to the dashboard widget.
• the graph method of Q effectively tells the widget to render a graph. If you want and bar-chart, area-chart... you can use Q.bar, Q.aread and so on.

The standard packages allow you to go beyond the default Qwiery capabilities. For example, this script reads in some csv data :

import csv
from js import XMLHttpRequest
from io import StringIO

req = XMLHttpRequest.new()
req.open("GET", "/drugs.csv", False)
req.send(None)


data = StringIO(req.response)
reader = csv.DictReader(data)
data_list = list(reader)

Q.grid(data_list)

See the playground for various (advanced) examples.

Playground ​

There is a Python playground available in the dashboard. You can use this to experiment with Python code.