For a researcher: your own data#

You have a folder, or several, that grew over years: dated experiment folders such as 20261012_HeLa_DAPI-GFP, files saved again with _final on the end, scripts that wrote tables that fed figures. You know roughly what is in it. You may want to know exactly: which files belong to which experiment, which were made from which, what repeats, and what is missing.

That is what Ingence reads out of a folder. It works from the names, folders, sizes and dates of your files, and, if you let it, from inside the files themselves. It does not interpret your science.

What you get#

Your work by time. The Dashboard opens on your experiments laid out by date. Each mark is an experiment, and its color says how much of its raw data was followed by an analysis.
Experiments with no analysis found. A list of experiments that have raw data and little or no analysis among the files Ingence read: at least five raw acquisitions found, fewer than half followed by an analysis, and fewer than five analysis files and three results tables of the experiment’s own. It says “found”, not “exist”: the analysis may be on another drive. You can say where it is, and the next run leaves it off the list.
What repeats. Files that are byte-for-byte copies, replicates of one condition, and files that look made from others.
A reason for every conclusion. Open any file and see what was read from it and each conclusion about it. Each names the rule that made it, and the rule is written in plain words on the Engine page (How Ingence thinks).
Questions, not guesses. Where a rule cannot decide, Ingence asks, and nothing changes until you answer.

What it needs from you#

Your folder uses your words, and only you know what they mean. The words nothing reads are listed on the Coverage part of the Questions page, and only some of them are raised as questions, the most useful to answer first. Your answer is kept in a log under your name and never overwrites an earlier one (Teaching Ingence your words). What Ingence concludes waits for Re-run, so the conclusions on the screen do not shift while you read them. A question you answer leaves the list of open questions at once, and a name you give a line of work shows at once.

If your field has words of its own, a domain pack adds them in one step. If you work in a laboratory whose words are not in a shipped pack, you can write one.

Starting on your own folders#

Put the project somewhere else. The project is where Ingence keeps what it finds; keep it on your own disk, outside the folder it reads (Privacy and your data).
Start with the Dashboard. Typing ingence see with nothing after it opens a page that makes a project, asks the little the engine needs and reads the folder. To do the same steps one command at a time, follow Your first project, and use your folder as the root.
Read it at its own pace. On a large folder, ingence index -p PROJECT --dry-run --seconds 60 shows what a census would find without writing anything.
Keep private folders out. To skip a folder altogether, add its name to scan.exclude_globs in the project’s settings file. A list you write there replaces the default list, so start from the default shown in the settings reference.

What it will not do#

It will not move, rename, delete or change a file of yours. It reads, and writes only into its project and a small settings folder for your user account.
It will not tidy your folders for you. It tells you what it sees.
It will not send your files or their names anywhere, and it needs no account.
It will not tell you whether your experiment worked. It describes how your files relate.

Where to go next#

Your first project takes ten minutes on a practice folder. How Ingence thinks explains the ideas behind what you see.