Best Practices
Practical habits that make uploads, search, maps, 3D scenes, and predictions easier to review.
Overview
Mineflow works best when the original exploration context stays close to the data. The recommendations below are about making the product easier to audit: complete uploads, clear provenance, early map checks, and careful review before exporting or using prediction outputs.
Workflow Habits
Upload the project folder when you can, not only the cleaned spreadsheet. Reports, maps, sidecars, tables, and filenames often provide the context needed to classify and join records correctly.
Shapefiles, MapInfo, ENVI, File Geodatabases, ADF grids, Zonge packages, and many raster formats need companion files. Upload the full bundle together.
After indexing, open the data on the map and verify that collars, layers, rasters, and polygons land where you expect. Projection or sidecar problems are easier to catch visually.
Use Search Scope to decide whether a search should use public data, personal uploads, or both. Use Data Sources when you want to narrow the public side by provenance.
Use Documents for source-level review, list search for text triage, and Map Search when the answer depends on location, geometry, or relationships between layers.
Before relying on a prediction, review the target, selected model, parameters, and input coverage. Compare the output against the original assays, lithology, maps, and geological context.
Related Pages
- Data Uploads explains how to add sample datasets, files, folders, and cloud links.
- Supported File Formats lists supported files and sidecar bundles.
- Search Overview explains Search Scope, Data Groupings, and Data Sources.