---
title: "How importing data works in MangoMap"
description: "What happens after you click Save on an import, why MangoMap reprojects data to Web Mercator, and which column names it changes and how."
sidebar:
label: "How importing works"
---
When you import data, MangoMap stores it in your account library as a **dataset**, then makes it available to draw on any map as a **layer**. This page explains these terms, what happens after you click `Save`, and why MangoMap changes the projection.
## Datasets and layers
A **dataset** is a copy of your data inside your account library. It lives at `Data` in the main navigation, and it's independent of any particular map. A **layer** is that dataset placed on a specific map, carrying its own styling and visibility. One dataset can power layers on many maps. For the full picture — what belongs to the dataset versus the layer, and why removing a layer keeps the data — see [Datasets and layers in MangoMap](/docs/importing-data/concepts-datasets-and-layers).
This split matters when you import:
- Importing from the `Data` page adds the file to your library only. You attach it to maps later.
- Importing from a map (`Manage layers` → `Import new data`) does both at once — uploads the dataset _and_ adds it as a layer on that map.
## What happens after you click Save
Your files upload as soon as you add them. When you click `Save`, MangoMap processes them:
1. The page shows a `Processing data…` dialog while MangoMap prepares the dataset.
2. MangoMap reads your file and prepares the data.
3. On success, you land on the dataset page (or back on the map with the new layer added). On failure, the dialog switches to `Processing failed` with an error message and a `Try again` button. `Try again` clears the form, so you add the files again.
Processing time depends on the file size and complexity. Closing the tab does not cancel processing, and the dataset still appears in your library when it is ready. But the page can't show you the result, and if you started from a map, the layer is not added to it.
Follow an upload through these states to see where it can fail and where `Try again` takes you:
## Why MangoMap reprojects to Web Mercator
MangoMap converts your data to match the basemap, so you never pick a coordinate system. It reads the one your file declares:
- For a shapefile, that's the `.prj` file.
- For GeoJSON, the `crs` field if present, otherwise `EPSG:4326` per the spec.
- For other vector formats, whatever the file declares internally.
If MangoMap finds no coordinate system with a code, such as `EPSG:27700`, it assumes `EPSG:4326` (WGS84). If the file names a system MangoMap can't convert from, processing fails with an `Unsupported coordinate system` error and you'll need to reproject the file in another tool first. See [Unsupported coordinate system](/docs/importing-data/troubleshoot-unsupported-crs).
The diagram shows which system MangoMap converts from, and the one case that fails:
MangoMap does not keep the original coordinates with the Web Mercator copy. If you need to keep the source projection, keep your original file outside MangoMap.
## Reserved field names
These five column names are reserved for system use. If a column in your file has one of them, in any mix of upper and lower case, MangoMap renames it on import by adding a leading underscore:
| Source name | Stored as |
| --- | --- |
| `gid` | `_gid` |
| `geom` | `_geom` |
| `id` | `_id` |
| `geometry` | `_geometry` |
| `the_geom` | `_the_geom` |
This is the full list. It applies to every vector and tabular format (`CSV`, `GeoJSON`, `KML`, `Shapefile`, `TAB`, `Geo database`).
### CSV coordinate columns
A CSV has no geometry, so MangoMap makes each row's point from two columns. It looks for these names, in any mix of upper and lower case:
| Coordinate | Column names |
| --- | --- |
| Longitude | `Longitude`, `Lon`, `Long`, `Lng`, `X`, `Longtitude` |
| Latitude | `Latitude`, `Lat`, `Y` |
The first matching column from the left is used for each. Those two columns become the point's location and are not kept as attributes, so they do not appear in the data table, the popup, or the style field lists. Another column that also matches, such as `X` when `Longitude` is used, stays an ordinary attribute.
To keep the coordinates as attributes, add a copy of each column with another name, such as `lon_value` and `lat_value`, before you import. The values must be decimal degrees (WGS 84). See [How to import a CSV](/docs/importing-data/import-csv).
## Other column name changes
MangoMap also changes a column name that is not a valid database name. In this order, it:
1. Removes spaces at the start and end.
2. Replaces each run of spaces with one underscore: `Land use` becomes `Land_use`.
3. Removes every character that is not a letter, a digit, or an underscore: `Cost (€)` becomes `Cost_`.
4. Adds a leading underscore to a name that starts with a digit: `2020 population` becomes `_2020_population`.
5. Adds a leading underscore to a reserved name (see above).
For every format, the stored name is also lower case: `Shape_Leng` becomes `shape_leng`.
If two columns end up with the same name, ignoring case, the import stops with `Duplicate column names found after sanitization`, followed by the names. Rename those columns in your file so each name is unique, then import it again.
Follow one column name through every rule to see what MangoMap stores:
## Where you see the new name
MangoMap keeps the original name as the column's display name, so the style field lists and the legend still show `id`. The data table, popup variables (`{_id}`), downloads, and anything that reads the data from outside MangoMap use the stored name. If you reference the column from outside MangoMap, use the stored name.
This also matters when you upload a new file to update an existing dataset — see [What can break](/docs/data-updates/concepts-data-updates#what-can-break) in the data updates concept page.
## Common confusions
**"Why is my data showing up in the wrong place?"** — Usually a coordinate-system problem. The file declared the wrong projection, or had no projection information and MangoMap defaulted to WGS84. For shapefiles, check that the `.prj` was uploaded with the others. For GeoJSON, confirm the file is in WGS84.
**"I deleted the layer but the dataset is still there."** — That's expected. Layers and datasets are independent. To remove the data entirely, delete the dataset from the `Data` page.
**"I uploaded the same shapefile twice and got two datasets."** — MangoMap does not deduplicate by content. Each import produces a new dataset.
## Related
- [Supported file formats and limits](/docs/importing-data/supported-formats)
- [How to import a shapefile](/docs/importing-data/import-shapefile)
- [How to import a CSV](/docs/importing-data/import-csv)