Skip to main content

Export and Predict

After you've created your model, export it to use it in production.


You can download visualizations you've seen throughout the Machine Learning process:

  • Correlation Matrix

  • Chord Diagram

  • Confusion Matrix

  • Details about the Target

  • Feature Importance

Choose whether you want to download the visuals as a PowerPoint file (PPTX) or as a ZIP file that contains PNGs of each visual, then select Export Visuals.

Designer Integration

Use the Machine Learning Send and Machine Learning Predict tools to stream data into Machine Learning, make predictions, and then return it to your Designer Desktop workflows.

To use your model in Designer Desktop, export your model to a YXMD file and then open it in Designer Desktop. The workflow contains a Machine Learning Predict tool already on the canvas. To use the Machine Learning Predict tool...

  1. Go to the downloads page.

  2. Go to Alteryx Intelligence Suite > Alteryx Intelligence Suite (your Designer Desktop version).

    1. Note that you must have Designer Desktop 21.3 or higher installed.

  3. Download the Alteryx Machine Learning Designer Integration Tools YXI file.

  4. Open Designer with the "Run as Administrator" option enabled.

  5. Open the YXI file to start the installation process.

Python Script

Select Export Script to export your model as Python code. Export your model to evaluate the underlying code. Follow these steps to run your Python model code…

Software Requirements

  1. Python 3.8 to 3.10 installed. Download the latest version of Python here.

  2. An IDE for running Python. We recommend the free PyCharm community version. Download the latest version of PyCharm here.

  3. Install dependent libraries. The model export folder includes arequirements.txtfile that references the required libraries. PyCharm automatically infers this file and installs the required libraries. If you choose to use an IDE other than PyCharm, run the pip install -r requirements.txt command to install the required libraries.

How to Use the Python Model

  1. Extract the file. The ZIP file contains your training and holdout data, a library requirements TXT file, a PY file with executable model code, a column mapping JSON file, and a features JSON file.

  2. Open your IDE and reference the extracted modelExport folder in a new project.

  3. Run the code from your project to generate model component results.

  4. For more information on the underlying code, refer to the EvalML documentation.

Upload New Data for Predicting

To make model predictions for Regression and Classification models, select new data that Machine Learning hasn't seen before. You can Browse for a file from Alteryx Analytics Cloud or Import a file from your computer. The new data must include the columns, in any order, used to build the model. It's OK if the new data contains a column with the target variable.


We treat rows that start with a number sign (#) as comments. If the first row in your CSV file is a comment, then the second row becomes the column header.

We also remove empty rows after the header row in your CSV file. If rows contain some empty cells, we convert these empty cells to null values instead. Note that it's good practice to upload files without empty rows.

Time Series Forecast Graph

If you select a Time Series model, you can view your historical data and your predicted, or forecasted, data in a single graph. To change the predicted data dates in the graph, go to Auto Model > Time Series Setup > Intervals.

The forecast graph also includes prediction intervals at 85% and 95% confidence levels. Use prediction intervals to determine the confidence of a forecasted data point.