.. _KNIME: https://www.knime.com/ .. _Python: https://www.python.org/ .. _pandas: https://pandas.pydata.org/

==================================== knime(py): Python toolkit for KNIME

knime(py) provides tools for reading and executing KNIME_ workflows from Python_. It is distributable as a single file module and has no requirements beyond Python_ 3.6+ and the Python Standard Library <http://docs.python.org/library/>. Optionally, if the pandas module is also installed, then pandas DataFrames are supported for both input and output to KNIME workflows executed through this toolkit.

Example: Execute a KNIME Workflow

.. code-block:: python

import knime

with knime.Workflow("DemoWorkflow01") as wf: wf.execute() results = wf.data_table_outputs[:]

Download and Install

.. __: https://github.com/KNIME/knimepy/blob/master/knime.py

Install the latest stable release with pip install knime (or pip3 install knime if you have both Python 2 and 3 installed). Alternatively, download knime.py__ (unstable) into your project directory. There are no hard dependencies other than Python 3.6+ and the Python standard library itself.

License

.. __: https://github.com/KNIME/knimepy/blob/master/LICENSE

Code and documentation are available according to the LICENSE__.

Expanded Example: Multiple Inputs/Outputs when Executing a KNIME Workflow

.. code-block:: python

import knime import pandas as pd

Change the executable_path to point at a particular KNIME install.

May alternatively be set via OS Environment Variable, 'KNIME_EXEC'.

knime.executable_path = r"C:\Program Files\KNIME\knime.exe"

Prepare input data tables as DataFrames or regular dicts (in KNIME's

required schema) to be read by the "Container Input (Table)" nodes in

the KNIME Workflow.

input_table_1 = pd.DataFrame([["blau", -273.15], ["gelb", 100.0]], columns=["color", "temp"]) input_table_2 = { "table-spec": [{"color": "string"}, {"size": "long"}], "table-data": [["blue", 42], ["yellow", 8675309]] }

Use a with-statement to set the inputs, execute, and get the results.

with knime.Workflow(r"C:\Users\berthold\knime-workspace\ExploreData01") as wf: wf.data_table_inputs[0] = input_table_1 wf.data_table_inputs[1] = input_table_2 wf.execute() output_table = wf.data_table_outputs[0] # output_table will be a pd.DataFrame

KNIME Server Example: Execute a KNIME Workflow on a remote KNIME Server

.. code-block:: python

import knime import pandas as pd

input_table_1 = pd.DataFrame([["blau", -273.15], ["gelb", 100.0]], columns=["color", "temp"])

Requires a valid user account on a running KNIME Server instance.

with knime.Workflow( "https://your_server.your_company.com/knime/your_directory/your_workflow_name", username="arthur.dent", password="forty-two" ) as wf: wf.data_table_inputs[:] = [input_table_1] wf.execute(reset=True, timeout_ms=10000) # Default timeout is usually plenty. output_table = wf.data_table_outputs[0]

Alternative syntax for specifying the Workflow, keeping everything else the same:

with knime.Workflow( workspace_path="https://your_server.your_company.com/knime", workflow_path="/your_directory/your_workflow_name", username="arthur.dent", password="forty-two" ) as wf: wf.data_table_inputs[:] = [input_table_1] wf.execute(reset=True, timeout_ms=10000) output_table = wf.data_table_outputs[0]