Python Scripts

Python is an interpreted high-level programming language for general-purpose programming. Python features a dynamic type system and automatic memory management. It supports multiple programming paradigms, including object-oriented, imperative, functional and procedural, and has a large and comprehensive standard library.

A few prerequisite steps need to be taken in order to integrate Python 3 scripts into Iotellect, indicated below.

Download and install Python in your system. It is recommended to install Python as a system-wide component rather than by user. There are a few ways to install python:

• Python source documentation offers installation instructions for a wide variety of operating systems

• Python can also be installed via Anaconda, which may simplify the installation of the JEP dependency.

Python Versions Current version of Iotellect is compatible with versions of Python 3.5 or higher. Some older versions of Iotellect are compatible only with Python version 3.6.

Install the latest Visual C++ Redistributable all-in-one package for 2015-2022 versions.

Several environment variables should be set or updated. The following changes should be made for the user Iotellect will run under. That is, if Iotellect runs as a service on Windows, system variables should be updated, rather than the user level variables.

• PYTHONHOME - root directory of your Python installation

• Add PATH to paths: %PYTHONHOME% (or $PYTHONHOME) and %PYTHONHOME%/Scripts (or $PYTHONHOME/Scripts)

Pip is the package manager for Python, which can be installed with the following:

# or download from a browser if you haven't curl
curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
python get-pip.py 

The integration requires the open source Pandas data analysis library, which adds a number of useful tools to the standard python library. In most systems, the package can be installed as follows:

Pip

pip install pandas

Anaconda

Pandas and its dependence are usually included by default in new Python environments created with Anaconda. Pandas can be installed with the following if needed:

conda install -c conda-forge pandas

JEP provides a way to execute Python code from within Java applications and enables access to Python modules, objects, and functions as if they were Java objects. This functionality allows Iotellect to interact with Python scripts.

Windows Users: It is recommended that you install JEP from source or unpack an appropriate jep-distro archive to the location where Pandas was installed. To identify the path, run python show pandas and find the Location label.

Linux Users: Simply install JEP with pip or Anaconda: Pip pip install jep Anaconda conda install -c conda-forge jep

Might be necessary only for older Iotellect versions. Skip this step if using the latest version of Iotellect.

Copy the "jep" folder from Iotellect Server "lib" directory into Python's "site-packages" folder, if Python scripts will be used only for the current user.

• To find the location of the "site-packages" directory, use command: python -m site --user-site.

Copy the "jep" folder from Iotellect Server "lib" directory into Python's "dist-packages" folder, if Python scripts will be used for all users.

• To find the location of a "dist-packages" directory, use command: python -m site.

Reboot the machine and start Iotellect server.

Scripts are written in Python. Every script must contain:

dataSetDataFrameInput
dataSetDataFrameOutput

Where dataSetDataFrameInput is a pandas.DataFrame converted from the input Data Table.

• All Iotellect numeric data types (e.g. Integer, Double or Long) and Date type are converted to the Python double type.

• All Iotellect string and color types are converted to the Python character type (str).

• Every nested Data Table will be converted into a nested pandas.DataFrame.

During transformation of a Data Table into pandas.DataFrame, all column names remain the same. In the pandas.DataFrame, the number of rows and columns corresponds to the number of Data Table records and fields, respectively.

The result of execution must be assigned to the dataSetDataFrameOutput.

The "pandas" library is not included in the standard Python interpreter. The "pandas" library must be installed separately, as indicated above.

When a new script is created, its text is not empty. It contains an auto-generated code with two variables:

dataSetDataFrameInput #Input Parameters (pandas.DataFrame)
dataSetDataFrameOutput #Output Parameters (pandas.DataFrame)

To execute Python scripts your OS should have Python environment variables.

Python Script Example:

import pandas

dataSetDataFrameInput #scrScriptPyDefaultDataFrame
pythonList = list()
pythonList.append(1)
pythonList.append(2)
pythonList.append(3)

innerStrList = list()
innerStrList.append("newStr")
innerStrList.append("someStr")
innerStrList.append("superStr")
pythonList.append(innerStrList)

df = pandas.DataFrame(pythonList) 

dataSetDataFrameOutput = df #scrScriptPyDefaultDataFrame

Was this page helpful?

Tell us more

Send

Thank you!