Import data

This guide walks through a single import end to end: choose a data type, upload a file, review the per-row validation results, resolve or accept what Erde flags, and commit. Commit runs in the background, so the import moves from Committing to Committed on its own.

Before you start

• Role required: permission to add data to the target Project (or, for site-scoped types, the target Site). The Project or Site must also be unlocked.
• Prerequisites: the Site the import belongs to — and, for operational types (samples, lab data, monitoring), the Project — must already exist. See Manage sites and Create your first project.

For the file column layouts each type expects, see Import formats.

Steps

1. Go to Imports in the left navigation, then select the Add (+) button on the imports list. This opens New Import (/imports/new).

2. Select the tile for the data type you want to import — for example Sample, Lab Results, Water Level, or Logger. Tiles are grouped by family — Site (locations, coordinates, wells, boreholes, sediment cores, test pits), Analytical (samples, lab reports, lab results), and Monitoring (water levels, well purges, field parameters, loggers) — to help you find the type you need. Only types that have an active import format appear as tiles.

3. On the {Type} Import page, fill in Import Settings:

   Field	What to enter
   Site	The Site the data belongs to. Required.
   Import Format	The format to parse the file. Shown only when the type has more than one format.
   Project	The Project that owns the data. Shown and required only for operational types (samples, lab data, monitoring, loggers).
   Import Mode	Create. This is the only mode offered.
   Import Name	Optional label. Leave blank to auto-generate it from the file name.
   Description	Optional notes.

   For a Logger import, a Load Mode field appears (bulk or single stream); choosing single stream adds Location and Stream fields.

4. Under Data Source, leave Source set to File Upload (some formats also offer Manual Entry; this guide covers file upload). Optionally select Download Template to get a blank file with the expected columns. (A Logger import instead offers Wide template and Tall template — either layout is auto-detected on upload.) Then select Choose Files and pick your file. Erde accepts CSV (.csv, .txt) and Excel (.xlsx, .xls), and you can select more than one file (a Logger import takes exactly one).

5. Select Create Import. Erde lands you on the import detail page (/imports/{id}) with a status of Draft.

6. Select Validate. Validation runs in the background (just like commit): the import moves to Validating, the rows grid is replaced by a Validating import… notice, and Erde parses every row and tags each one Valid or Invalid, sets a planned Action (Create or Skip), and attaches any warnings. You do not need to keep the page open — Erde refreshes the status on its own and sends you a notification when validation finishes. When it completes, the grid returns and the status badge becomes one of:

   Status	Meaning	What to do
   Ready to Commit	Every row is valid and at least one row will create a record.	Continue to step 8.
   Validation Failed	One or more rows are Invalid.	Fix the flagged rows (step 7), then re-validate.
   Nothing to Commit	Validation passed but every row would be skipped or is excluded.	Edit the data or upload a different file.

7. Review and resolve the rows. Each entity gets its own tab, captioned {Entity} ({row count}). The rows grid shows the row #, Status, planned Action, Source, your data columns, and a Feedback column with the validation message. For a lab import, the Analysis tab also shows a read-only summary under the grid of the batches commit will create from the typed batch columns — one line per distinct code and type — which recomputes as you edit rows and is never itself editable. The grid loads one page of rows at a time from the server, so even very large imports open quickly; use the page controls at the top of the grid to move through the data — First / Previous / Next / Last, a Go to page box, and the Rows per-page selector. To work through problems:

   • Turn on the Errors only filter to focus on Invalid rows. The filter is applied on the server, so it finds the error rows across the whole import, not just the current page.
   • To review the whole list at once — or hand it to whoever produced the file (for a lab report, the lab) — select Export Errors for a rolled-up Excel summary of every validation error, Export Warnings for the non-blocking concerns, or Export Information for the notes on what the import resolved or defaulted automatically. See Export errors, warnings, and information.
   • Select Filters to narrow the rows by any column. Pick a column, an operator (Equals, Does not equal, Contains, Is blank, or the numeric/date comparisons Greater than / Less than / or equal), and a value; add more conditions with Add condition (all conditions must match), then Apply. Like Errors only, the filter runs on the server across the whole import. For a column with a fixed list of values (units, statuses, analytes, etc.), the value box offers one entry per value and matches a row whether it stored that value's name or its code. Comparisons cast the value first, so Greater than 100 compares numerically (not as text). Clear removes the conditions.
   • Edit a cell in place to correct a value. Each edit saves immediately — there is no separate Save step. A saved edit returns that row to Pending (its previous validation no longer applies) and the import to Draft, so re-validate afterward to recheck. Edited rows stay visible until you change page, re-apply a filter, or re-validate.
   • To fix the same wrong value everywhere, select Replace All: choose a column, the value to find (leave it empty to target blank cells), and the value to replace with (leave it empty to clear the cell). Preview matches counts the affected rows first. The replacement runs on the server across the rows the current filter is showing, so combine it with Errors only or a column filter to scope it. Replaced rows return to Pending — re-validate afterward.
   • To accept a row that cannot be imported, right-click it and select Exclude Row. Excluded rows are skipped and never committed; Restore Row brings one back. Right-click also offers Revert Row to Original, which discards that row's edits and restores its as-imported values.
   • To act on many rows at once, tick their checkboxes in the leftmost column — or the header checkbox to select the whole page — and use the Exclude selected or Revert selected button that appears in the toolbar. Exclude selected excludes every selected row (rows already excluded are left as-is); Revert selected discards edits and un-excludes the chosen rows, returning them to Pending. Both apply across the whole selection in one step.
   • Re-run Validate until the import reaches Ready to Commit.

8. Select Commit. Confirm in the Confirm Commit dialog, which states how many entities will be created. The import moves to Committing.

9. Wait for the commit to finish. While the commit runs, the rows grid is replaced by a Committing import… notice — the data is locked during the commit, so there is nothing to edit. Erde processes the commit in the background and refreshes the status automatically; you do not need to keep the page open or reload it. When it finishes, the grid returns and the status becomes Committed (success) or Commit Failed.

Figure: The import lifecycle from creation to a committed result.

Result

The import shows a Committed status badge, and its statistics strip reports Created (and, where applicable, Skipped) counts. Every record Erde created is stamped with this import, so the whole batch can be removed in one step if needed — see Troubleshoot and roll back imports.

Enter data manually

For most data types you can key rows in directly instead of uploading a file — useful for a handful of records. Manual entry is offered for every type except lab reports, lab results, and loggers (those are file-only).

1. Create the import as in the steps above, but under Data Source set Source to Manual Entry. No file is required.
2. Select Create Import. Erde opens the import in Draft with an empty row set.
3. On the import detail page, add rows by hand and fill in their cells.
4. Select Validate, resolve any flagged rows, then Commit — the rest of the flow is the same as a file import.

Find and manage imports

The imports list at Imports (/imports) is where you return to an existing import. It pages through every import server-side and supports:

• Site and Project filters, plus a status filter you can clear.
• A My imports view that shows only the imports you created.
• Search and export of the list.
• Server-side sorting and per-column filters (with comparison operators) across the grid columns.

Select any import to reopen its detail page and continue validating, committing, or rolling it back.

Troubleshooting

Symptom	Cause	Resolution
No tile for the type you want	That type has no active import format on this deployment.	Check available formats in Import formats, or ask an administrator.
Status is Validation Failed after validating	At least one row is Invalid.	Filter to Errors only, read each row's Feedback, fix the cells (edits save automatically), then re-validate.
Status is Nothing to Commit	Validation passed but every row would be skipped or is excluded.	Edit the data or upload a different file, then re-validate.
Commit button does not appear	The import is not in Ready to Commit.	Validate first; resolve any Invalid rows so the status reaches Ready to Commit.
Status stays Validating	The background validation is still running; large files take longer.	Wait — the page updates the status on its own and notifies you when validation completes.
Status stays Committing	The background commit is still running; large files (such as logger readings) take longer.	Wait — the page updates the status on its own when the commit completes.
Status is Commit Failed	The commit could not finish.	Open the import, read the error, edit and re-validate the rows, then commit again.

Related

• Import formats
• Troubleshoot and roll back imports
• The import pipeline