Welcome to
We provide you with a Python package that delivers the Command Line Interface (CLI), Malevich Square, and Malevich Meta services to your system. These services are used to build flows and apps, that will then appear in your workspace.
The package is distributed using pip package manager. Run the following command to install Malevich to your system:
python3 -m pip install malevichCheck the setup with
malevich --helpTo first cool thing you can do with Malevich library is deploying your flow to UI! To do it, you need firstly tell the library who you are. This is done using a file myspace.yaml which stores your credentials. Use the following command to login to Space:
malevich space loginYou will be prompted for username, password and optionally for Malevich Space URL and organization id.
When you install an app, you register it in your system using the malevich.yaml file and creating a stub. A stub is a generated script, that enables you to import Malevich apps as they are sub-packages of malevich. You can see, what apps are registered and available as stubs using
malevich listLet us start with installing an app called Utility. Utility is the most basic yet powerful app provided with Malevich. It helps you to glue flows and perform operations on dataframes. To install the app, use the following command:
malevich install utilityOnce you installed the app, it should appear as a subpackage. which means you may import a processor from it and explore it:
from malevich.utility import locsThe simplest possible flow you may write is selecting a collumn from a collection.
# myflow.py
import pandas as pd
from malevich.utility import locs
from malevich import flow, config, collection
from malevich.interpreter.space import SpaceInterpreter
@flow(reverse_id="my-first-flow", name="My First Flow")
def my_first_flow():
"""This function is a flow definition!"""
# Create a collection
my_collection = collection(
df=pd.DataFrame({'left_column': ['A', 'a'], 'right_column': ['B', 'b']})
name="My First Meta Collection"
)
# Select `right_column`
return locs(my_collection, config(column='right_column'))
# The function returns a deployment (a.k.a task), which
# can run on Malevich Space, or Malevich Core
deployment = my_first_flow()
# Deliver flow to the platform!
deployment.interpret(SpaceInterpreter())After running python3 myflow.py you will see a new flow named My First Flow in Malevich Space workspace
Check you the documentation to get an idea about other apps and possibilities of Malevich!