Gaplad › Free tools & guides › Missing timestamp finder

Free tool · runs in your browser

Find every missing timestamp in your time series.

Drop a logger export and see where the record has holes: which time steps are missing altogether, which sensors dropped out on their own, how long each gap is, and how complete each column is.

Drop a CSV or Excel file here

Read in this browser tab. Nothing is uploaded. One timestamp column (or Year/Month/Day columns) and any number of value columns.

Paste data instead

Why count gaps before you fill them

Most gap-filling advice starts with the method. The honest place to start is the gap list. A record with three hundred one-hour dropouts needs a very different treatment from one with a single three-week outage, even when both are "95 % complete". Short gaps of a few hours are filled well by linear interpolation; long gaps need a second sensor that kept recording; a gap where every sensor died together often cannot be filled honestly at all.

The length table above uses the same classes as Gaplad's validation (≤6 h, 6–24 h, 1–7 d, 7–30 d, >30 d), so you can see at a glance which kind of problem you have.

What it reads

Questions

How does it find missing timestamps?

It reads your timestamps, works out the logging interval (the most common step between readings), lays an unbroken grid from the first to the last reading at that interval and lists every grid step with no reading. Duplicate timestamps are counted, and readings a little off the grid snap to the nearest step.

What is the difference between a missing time step and a gap in one column?

A missing time step has no value in any column: the logger recorded nothing, or the row is absent from the file. A gap in one column means that sensor failed while others kept recording, which is the case where a gap can often be filled well from the sensors that survived.

Can it put the missing rows back into my file?

Yes. The second download writes your record on a complete regular grid with a row for every expected time step; the missing rows are inserted with empty cells. Nothing is filled or invented.

How do I do the same in Python or Excel?

In pandas: df.set_index('timestamp').asfreq('1h') inserts the missing rows, and df['v'].isna() marks them. In Excel, a helper column with =A3-A2 shows every step that is longer than the interval. This page does both without code and also lists each gap's start, end and length.

Is my data uploaded?

No. The file is read by JavaScript in this tab and never leaves your computer. You can disconnect from the internet after the page has loaded and it still works.

Found the gaps? Fill them, with proof

Gaplad fills each gap with whichever method wins on your own data: it carves synthetic gaps out of what you observed, fills them blind and scores every method against the truth before it touches a real gap. Every filled value is flagged in the export.

Open Gaplad, free How to choose a method