The hardest part of mapping data is rarely the map - it is the data. A spreadsheet that looks fine in cells can produce points in the ocean, missing markers, or an upload that fails outright. Here is a checklist to clean your data first.

Fix swapped or out-of-range coordinates

Latitude runs from -90 to 90; longitude runs from -180 to 180. If you see a "latitude" of 120, the columns are probably swapped. A point that lands off the coast of West Africa (0, 0) usually means a row had empty or zero coordinates - the infamous "Null Island."

Watch out for text in number columns

Coordinates stored as text - with stray spaces, degree symbols, or comma decimal separators like 51,50 instead of 51.50 - will not parse as numbers. Convert these columns to plain decimal numbers before exporting your CSV.

Standardize your headers

MapDataHub auto-detects common column names like lat, latitude, lng, lon, and longitude. If your columns are named something unusual like Y_COORD or Pos_X, rename them so they are recognized automatically.

Common problems to scan for

  • Empty rows or trailing blank lines at the bottom of the file.
  • Duplicate header rows pasted into the middle of the data.
  • Mixed coordinate systems - make sure everything is in WGS 84 (EPSG:4326).
  • Non-UTF-8 encoding, which garbles accented characters and place names.
  • Thousands separators in coordinates (e.g. 1,234.56), which break number parsing.

Let the uploader catch the rest

You do not have to get every row perfect. When MapDataHub converts your CSV, it validates each row and automatically skips entries with invalid geometry, so a few bad rows will not block the whole upload. Clean the obvious problems, then upload and review what landed on the map.

After uploading, click a few scattered points to verify their properties match what you expect. Outliers far from the rest of your data are the quickest way to spot a coordinate that slipped through.