Overview
DORA 2.0 experiments result in a set of data files that can be downloaded and analyzed. The .zip file contains all relevant output graphs, learning points, Feature Importance (SHAP) graphs, and summary files, allowing you to further analyze the data.
Steps to Export your Data
Once your Prediction Map is complete and Target Groups have been generated, you can export your results in two ways:
Export from the Prediction Map
Complete Step 4: Select Input Features
Finish the final step in your experiment.
Create Target Groups
To export your DORA 2.0 Prediction Map, you must first Create Target Groups.
This generates additional Feature Importance labels for each target group, which are included in the export.
Export Results Files
Once you have created target groups, click the Prediction Results icon in the top nav to open the Prediction Results panel:
Click the Export icon in the upper-right corner of the Prediction Results panel, then select the files you want to download:
To export everything, select Full Results to download a
.zipfile containing all outputsWhen the
.zipfile is ready, you've see a notification on your screen. Click Export to download:
Export from the DORA Main Page
Find Your Experiment
On the DORA main page, locate the experiment you want to download.
It should be at the Target Generation status.
Download Results Files
Click the three-dot action menu next to the experiment.
Select Download Results from the menu.
In a few seconds, you'll see a notification in the lower right of your screen. Click Export to download a
.zipfile containing your results:
Note: You can only download your full results from this page. If you only want VPS or Feature Importance files, you can download them from the Prediction Map.
Exported Data Contents
Note: The output files represent the Prediction Map results as they were at the time of download. If any changes are made to the Map afterward, the updated results can be exported again to capture those modifications.
To prevent edits to your Prediction Map, you can lock an experiment.
Prediction Map text file:
Experiment
Model performance QA/QC Output graphs:
Confusion_Matrix
R2_Valid
Model-wide and target-specific value plots and CSV SHAP analysis:
Model_SHAP
shap_summary_target_Beta
shap_summary_target_Gamma
shap_values_target_Beta
shap_values_target_Gamma
VPS results CSV (importable into GIS/3D modelling software):
resultsVPS
2D Raster file of VPS results (can be visualized in GIS software):
vps_results
vps_results.ers
Using Exported Feature Importance and VPS Data
Once exported, your results include Feature Importance (SHAP) summary files and VPS rasters, which can be used in your own GIS or 3D modeling software.
Feature Importance (SHAP Labels)
Use the Feature Importance files to explore which features are driving each modelled target. When imported into your software, you can visualize by target name to see how AI-identified areas align with geological expectations. This allows geoscientists to dig into the feature importance for each target, bridging AI insights with domain expertise.
VPS Rasters & CSV
VPS values are generated per pixel. To focus only on high-confidence results, apply a VPS threshold filter (e.g., show only values above 0.7).
Adjust layer transparency or symbology to highlight these zones without obscuring underlying geological layers. This helps replicate your DORA view and maintain clarity in your own environment.
💡 Tip: Watch for edge effects or anomalies in raster visuals. These may indicate data misalignment or overly influential layers and should be reviewed carefully before decision-making.
Learn More
Prediction Results:
Still Have Questions?
Reach out to your dedicated DORA contact or email support@VRIFY.com for more information.