How to use diagnostics after FIT ingestion

Once FIT ingest finishes, the next job is not to start tuning controls immediately.

The first job is to decide whether the dataset itself looks believable enough to continue calibrating from.

This article walks through the main diagnostics that answer that question.

What these diagnostics are for

After ingest, you are trying to answer:

Does this dataset look representative and strong enough to keep going, or should I improve the FIT folder first?

You are not looking for perfect charts.

You are looking for evidence that:

• the dataset is large enough to matter
• the terrain coverage is relevant to your real use
• heart-rate coverage is good enough to support calibration
• the overall relationships between grade, speed, and heart rate look believable

Step 1: Start with the FIT File Ingest Summary

Always begin with the FIT File Ingest Summary.

This is the shortest path to answering whether the dataset is viable at all.

This summary is the first screen to trust. It tells you whether the ingest produced enough usable activity and heart-rate coverage to justify looking deeper.

Pay the most attention to:

• Files analyzed
• the overall amount of usable activity
• whether heart-rate coverage looks strong enough to support calibration
• whether the altitude context looks like the kind of trips you care about
• whether the accepted usable data looks thin or substantial

Use this section to ask:

• do I have enough usable activity to say something real?
• do I have enough heart-rate coverage for calibration to be meaningful?
• does the altitude context look like the places I actually care about planning for?

If these top-level numbers look weak, it is usually better to fix the dataset before spending time on the later charts.

Step 2: Read Grade (%) vs speed before anything else shape-related

The Grade (%) vs speed (mph) chart is the first relationship chart to read.

It shows how observed moving-window speed behaves across grade.

This chart helps you judge whether the observed speed story looks coherent across uphill, flat, and downhill terrain. You are looking for a believable pattern, not a perfectly smooth cloud.

What you want to see:

• speeds that generally fall as uphill grade becomes steeper
• a believable center band rather than pure scatter
• enough spread across grade to make shape fitting worthwhile
• no obvious dominance by bizarre outlier behavior

Use it to ask:

• do uphill and downhill behavior look directionally reasonable?
• is the observed range broad enough to support a meaningful speed-shape fit?
• does the cloud look like hiking or backpacking data rather than unrelated activity noise?

If this chart looks chaotic, narrow, or unrepresentative, later calibration work will usually be weaker too.

Step 3: Use Grade (%) vs Heart Rate to judge strain by terrain

The Grade (%) vs. Heart Rate (bpm) chart helps you see how cardiovascular strain changes across grade.

This chart is useful for checking whether HR response rises in a believable way as terrain gets harder, while still keeping a sensible trough around easier grades.

What to look for:

• a generally believable rise in HR as uphill grades steepen
• a lower-strain trough near easier grades
• enough structure that ΔHR and effort-intent diagnostics later on will have something real to work from
• no obvious sign that HR data is too sparse or too erratic to trust

Use this chart to ask:

• does the HR story across grade look physiologically believable?
• does the dataset appear rich enough to support later ΔHR review?
• are there big coverage holes that make the curve hard to trust?

You do not need a perfect shape here. You need a believable one.

Step 4: Use Speed vs Heart Rate as the final sanity check

The Speed (mph) vs. Heart Rate (bpm) chart is a cross-check on the overall movement and strain story.

This plot helps you confirm that the dataset tells a coherent story about how speed and heart rate move together. It is a strong final sanity check after the grade-based charts already look believable.

Use it to ask:

• does speed generally rise into a sensible HR range rather than behaving randomly?
• do the bands look like real effort structure rather than mixed activity noise?
• does the overall cloud support the idea that this is a usable calibration dataset?

This chart is especially helpful when the grade-based charts look borderline and you want one more test of whether the data feels coherent.

Step 5: Check the altitude distribution last

The Altitude (ft) distribution chart appears lower on the page, so it works best as the final context check after you have already reviewed the main movement and HR relationships.

Use this chart to understand the altitude environment your dataset mostly reflects. In this example, most eligible movement sits in a fairly high-elevation band rather than being spread evenly from low altitude upward.

What to look for:

• where most of your valid samples cluster
• whether the median altitude looks plausible for your intended trip style
• whether the spread is broad enough to reflect your real use
• whether there are obvious outliers that are not representative

Decision rule:

• if the altitude band lines up with the trips you care about, that is a good sign
• if the dataset is dominated by a very different altitude regime from your intended planning use, be cautious about over-trusting the result

This does not automatically invalidate the calibration, but it tells you what kind of world the dataset is really describing.

A safe reading order

For most users, the safest order is:

1. Read the FIT File Ingest Summary.
2. Review Grade (%) vs speed.
3. Review Grade (%) vs. Heart Rate.
4. Use Speed vs. Heart Rate as a final sanity check.
5. Check the altitude distribution as the final context check.

That sequence follows the page more naturally and goes from “Is the dataset viable at all?” to “Do the relationships inside it look believable?” to “What altitude context does this dataset mostly reflect?”

Signs the dataset is probably good enough to continue

Good enough usually looks like:

• enough usable activity to support a real pattern
• usable heart-rate coverage
• an altitude regime that roughly matches your intended use
• grade-speed and grade-HR relationships that look coherent
• no obvious sign that the folder is dominated by irrelevant activities

You are looking for a believable dataset, not a pristine one.

Signs you should improve the dataset first

Be cautious if:

• the usable activity looks thin
• heart-rate coverage is weak
• the altitude band is clearly unlike the trips you care about
• the grade-speed relationship looks too noisy or too narrow
• the HR plots look sparse, erratic, or physiologically implausible

In those cases, improving the FIT folder is usually more productive than trying to force the calibration through later tabs.

What to do next

If the diagnostics look believable, continue with:

• How to use the calibration workspace tabs
• How to use the TRIPSpeed Shape tab and controls
• Review calibration outputs

If the diagnostics look weak, go back to:

• Prepare a FIT folder for calibration
• Which FIT files should I include for calibration?