ArchaeoDash was developed in R Shiny to facilitate the analysis of elemental data (for example portable X-ray fluorescence (PXRF) and Neutron Activation Analysis (NAA)). It includes tools for data management, ordination, clustering, distance-based matching, and group assignment.
This app was originally developed by Matthew Peeples and Andrew Upton and has been extensively redesigned by Robert Bischoff in consultation with Jeffrey Ferguson.
We recommend creating an account to reduce data loss risk. If the app times out, reloads, or encounters an error while not logged in, temporary work may be lost.
Click LOGIN in the side panel, check Register as a new user, and complete the form. You must accept the notice to complete registration. If you forget your password or username, email rbischoff@asu.edu.
Click LOGIN and enter your username and password.
Click BROWSE to upload data. CSV and XLSX are preferred, and other formats supported by the rio package in R may also work.
If you are logged in:
If you are not logged in:
Use Manage datasets to delete selected datasets or merge/rename selected datasets.
Click Press to confirm selections to apply these choices. Most analyses require this step before they update. A loading indicator is shown while the update runs.
When you confirm selections, you are prompted to name the transformation:
Each saved transformation can include:
Transformation behavior:
If Run LDA is selected with fewer than three groups, the app shows a warning because LDA visualization requires at least three groups.
Use the export section to download current data products. In most cases, confirm selections first so exported tables reflect current settings.
This tab has options for exploring the selected data.
This tab contains a plot displaying the number of missing or zero values for each element column as well as a view of the entire selected dataset. Data can be edited here by double clicking a cell and then clicking off the cell when the data has been changed. Extensive data modification should be done in a spreadsheet editor such as Excel and then uploaded to ArchaeoDash.
When logged in, edits autosave directly to the currently selected dataset in the database.
Warning: editing metadata columns (for example group or descriptive columns) changes database records. Make sure you maintain backups before making bulk edits.
Choose the columns you wish to group data by in column 1 and column2. Select the same column in both drop-down lists if you want to view results by only one column. Then select the type of display in the last drop-down. No result or an error may be returned if the data is not the correct type.
This tab shows histograms of the element data.
This line plot shows the variation in element concentrations across the dataset.
Use this tab to visualize data (elements, PCA, UMAP, LDA) and reassign selected points.
In visualize and select:
In multiplots:
Shows PCA outputs for the current confirmed transformation.
Shows LDA outputs for the current confirmed transformation. LDA requires at least three groups.
Cluster methods available:
Cluster analysis can be run using:
After running a method, use Record cluster assignments to add cluster labels back into the dataset. You can choose the output column name. If the column exists, you will be prompted to overwrite it.
This tab includes group size summaries and a full membership probabilities table.
elements,
principal components, UMAP, or
linear discriminants).The membership table supports filtering and column visibility control.
You can update group assignments directly from selected table rows:
BestGroup.These updates write back to the dataset and carry into downstream analyses and saved transformations.
Use this tab to identify closest matches by Euclidean distance.
Workflow:
elements,
principal components, UMAP, or
linear discriminants).The results table supports reassignment: