This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
Python Tool Doesn't Show Any Results or Errors on Run
When running the Python tool in Alteryx, the output is blank, and no results or errors are shown.
Python may have been previously installed by another app (e.g., Anaconda)
Product - Alteryx Designer
To ensure the issue is with the Python tool in Alteryx, create a simple 2-tool workflow with test data and code as below. Click Run in the Python tool and verify the tool returns no output and no errors.
Step 1 - Run the workflow with a Text Input Tool and Python code as below
Step 2 - Verify that the Alteryx engine reports no errors
Step 3 - Verify that the Browse tool shows no data available
The IPython kernel being called by the Jupyter Notebook may be referencing a previously installed or non-existent Python binary, rather than the Python binary installed with Alteryx.
Open an Administrator Command Prompt and navigate to:
Admin Version of Alteryx: C:\Program Files\Alteryx\bin\Miniconda3\PythonTool_venv\Scripts
Non-Admin Version of Alteryx: %LocalAppData%\Alteryx\bin\Miniconda3\PythonTool_venv\Scripts
Enter the command: jupyter kernelspec list
Navigate to the directory returned in step (2) in a File Explorer. If you have multiple directories, repeat steps 3-4 for each directory.
Open kernel.json in a Text Editor and verify that the first item in the argv list references the python.exe in the same directory as the directory from (1) or simply "python" in cases where the directory from (1) is included in your PATH Environmental Variable. To test if you have a valid argument, you can simply try entering the argument in a command prompt and it should open a Python shell.
When your Python libraries don't work the way they should in Python tool, restoring the tool to it's original state could be the solution. This article walks through how to restore Python libraries and the virtual environment associated with the Python tool.
Alteryx Designer comes with tools (based on both R and Python) to create and use predictive models without needing to write any code. But what if you've got custom models written in R or Python outside of Designer that you want to use in Designer, or vice versa?
If you are using a version of Designer ≥ 2018.3, use the Python tool! If you are on an older version, this article is for you.
Below is an example demonstrating the use of the Run Command tool to execute a Python script in the Designer and use its output in the workflow:
For a Python script, your Command should be Python.exe. If the directory where Python exists is in your system path variable, you can simply type Python.exe. Otherwise you will have to give it the full path, keeping in mind to quote the string if there are spaces (e.g. "Program Files"). In Command Arguments, you will type the location of your Python script (Alteryx's default working directory is the directory of the running module so it may be easiest to keep your script in the same folder to simply type "your_python_script.py" rather than the full path) and any - or -- options necessary. Remember to quote this string!
The example script from the attached simply sends text to a text file to be used as input for the Designer (note: it will not run unless you have configured your python environment 😞
As you can see, we’ve successfully executed a python script and used the Read Results input to bring the result of the script into the Designer for further processing.
The Python tool interface is based on a Jupyter Notebook. As of Alteryx Designer 2018.4, the Python tool can create logs for its web configuration. The frequency of these logs can be increased by switching on the debugging log level. Note that output file will be bigger and that it will impact Python tool performance in Designer.
With the Python Tool, Alteryx can manipulate your data using everyone’s favorite programming language - Python! Included with the tool are a few of pre-built libraries that extend past even the native Python download. This allows you to extend your data manipulation even further than one could ever imagine. The libraries installed are listed here - and below I’ll go into a bit more detail on what and why these libraries are so useful.
Each library is well documented, and there’s usually an introduction or examples on their sites to get you started on how a basic function in their library works.
ayx – Alteryx API – simply enough, we’re using Alteryx, sooo yea, kind of a requirement for the translation between Alteryx and Python.
jupyter – Jupyter metapackage – If you’ve used a Jupyter notebook in the past, you’ll notice the interface for the Python Tool is similar. This interface allows you to run sections of code outside of actually running the workflow, which makes understanding and testing your data that much easier.
matplotlib – Python plotting package – Any charting, plotting, or graphical needs you would want will be in this package. This provides a great deal of flexibility for whatever you want to visualize.
numPy – NumPy, array processing for numbers, strings, records, and objects – Native Python processes data in what some would call a cumbersome way. For instance, if you wanted to make a matrix, a.k.a. a 4x4 table, you would need to create a list within a list, which can slow processing a bit. However, NumPy has its own “array” type that fits the data in this matrix pattern that allows for faster processing. Additionally, it has a bunch of methods of handling numbers, strings, and objects that make processing a whole lot easier and a whole lot faster.
pandas – Powerful data structures for data analysis, time series, and statistics – This is your staple for handling data within Alteryx. Those who have used Python, but never pandas, will enter a whole new beautiful world of data handling and structure. Data manipulation within Python is faster, cleaner, and easier to code with. The best part about it is that the Python Tool will read in your Alteryx data as a pandas data frame! Understanding this library should be one of the first things to know when tackling the Python code.
requests – Python HTTP for Humans – for all the connector/Download Tool fans out there. If any of you are familiar with making HTTP requests (API calls and the like), then you should introduce yourselves to this package and explore how Python performs these requests.
scikit-learn – a set of Python modules for machine learning and data mining – Welcome to the world of machine learning in Python! This library is your go-to for statistical and predictive modeling and evaluation. Any crazy and wild methods you’ve learned for machine learning will most likely be found here and can really push the boundaries of data science.
scipy – Scientific Library for Python – all your scientific and technical computing can be found here. This library builds off the packages already installed here, like numPy, pandas, and matplotlib. Dealing with mathematical models and formulae are usually located within this library and can help provide that higher level analysis of your data.
six – Python 2 and 3 compatibility utilities – For those who are unfamiliar, Python versions come in 2 forms, version 2.x and 3.x (with 3.x being the most recent). Now, even though Python 3 is supposed to be the latest and greatest, there are still many users out there who prefer using Python 2. Therefore, integration between the two is a bit tricky with syntax differences, etc. The six module provides functions that are usable between the two so everyone can remain calm and happy! Their documentation is usually coupled with which version the functions most closely align to, so a user can get a better idea to its functionality.
SQLAlchemy – Database Abstraction Library – SQL in Python! Covers all your database needs from connecting to and extracting data, allowing it to interact with your Python code and thus, Alteryx itself.
statsmodels – statistical computations and models for Python – This library builds off sci-kit learn but focuses more on statistical tests and data exploration. Additionally, it utilizes R-style formulae with pandas data frames to fit models!
These are the libraries installed with the Python Tool, which can do almost any data function imaginable. Of course, if you’re looking to do something that these libraries don’t provide, there are myriad other Python libraries that I’m sure will help you with your use case. Most of these are also well documented in how to use so search away and let your mind float away in the beautiful cosmos created by Python.