Demo Data Hub

Data checks Records Reports Methods Analysis library

Data checks

Advisory questions about the generated records. Opening a question changes nothing.

Correct the record · 3

Executed · 2 questions on 268 records For review

Can this field-entered species label be resolved?

The entered label does not currently resolve to one accepted taxon label.

What to do: Map the entered name to its accepted plant in the name crosswalk.

Run 17 Aug 2026 Review 268 records

Executed · 6 questions, one per record Warning

Can this recorded patch relationship be repaired?

The source recorded a parent value, but it did not resolve to a generated patch key.

What to do: Match the visit to its mapped patch, or map the patch it names.

Run 17 Aug 2026 Review 6 records

Executed · 7 questions, one per record Warning

Should treated area be reconciled with mapped patch area?

The generated treated area is more than the configured tolerance above the generated mapped area.

What to do: Correct the treated area or remap the patch outline.

Run 17 Aug 2026 Review 7 records

Confirm in the field · 2

Executed · 9 questions, one per record Warning

Was treatment intended at the recorded green level?

A treatment method is present while the recorded green percentage is below the configured review threshold.

What to do: Ask the crew whether spraying at that green level was intended, then note the answer.

Run 17 Aug 2026 Review 9 records

Executed · 8 questions, one per record For review

Does the rapid regrowth sequence need review?

A generated follow-up records high cover soon after a complete-removal band.

What to do: Look again on the next visit; high cover so soon after full removal is unusual.

Run 17 Aug 2026 Review 8 records

Executed · 9 questions on 114 records For review

Plan around it

Is a follow-up visit due for this unit and species group?

The latest generated visit is beyond the configured re-treatment interval.

What to do: Add these patches to the next field round for their area and plant.

Run 17 Aug 2026 Review 114 records

What each check asks

Every check below ran on the generated records through 17 Aug 2026. Each one asks a question; none of them changes a record.

Correct the record

Can this field-entered species label be resolved?: The entered label does not currently resolve to one accepted taxon label.; What to do: Map the entered name to its accepted plant in the name crosswalk.; For review · Executed · 2 questions on 268 records
Can this recorded patch relationship be repaired?: The source recorded a parent value, but it did not resolve to a generated patch key.; What to do: Match the visit to its mapped patch, or map the patch it names.; Warning · Executed · 6 questions, one per record
Should treated area be reconciled with mapped patch area?: The generated treated area is more than the configured tolerance above the generated mapped area.; What to do: Correct the treated area or remap the patch outline.; Warning · Executed · 7 questions, one per record

Confirm in the field

Was treatment intended at the recorded green level?: A treatment method is present while the recorded green percentage is below the configured review threshold.; What to do: Ask the crew whether spraying at that green level was intended, then note the answer.; Warning · Executed · 9 questions, one per record
Does the rapid regrowth sequence need review?: A generated follow-up records high cover soon after a complete-removal band.; What to do: Look again on the next visit; high cover so soon after full removal is unusual.; For review · Executed · 8 questions, one per record

Plan around it

Is a follow-up visit due for this unit and species group?: The latest generated visit is beyond the configured re-treatment interval.; What to do: Add these patches to the next field round for their area and plant.; For review · Executed · 9 questions on 114 records

Thresholds in use Fixed in this demonstration

The values the checks and the Action center read. None can be changed here.

Check	Setting	Configured value
Action center: recheck overdue	Days after a full clearance	365 days
Action center: past the ceiling	Days since the latest visit	1,095 days
Is a follow-up visit due for this unit and species group?	Days since the latest visit	1,095 days
Should treated area be reconciled with mapped patch area?	Treated area allowed above the mapped area	10 %
Was treatment intended at the recorded green level?	Green cover below which a herbicide visit is reviewed	50 %
Does the rapid regrowth sequence need review?	Window after a full removal	14 days

Invasive Plants

Records

One record type at a time.

Demo data

Record type

Patches Visits Methods Plant names Form fields Not recorded Codebook

What the columns mean

Patch: The public alias of one mapped patch or point record.
Plant: The accepted plant name.
Record type: Mapped patch (an outline) or point record.
Field area: The field area, by its public label.
Mapped date: When the patch was first mapped.
Recorded area: Mapped area in hectares; a point record has none.
Visits: Visits on file for the patch.
Last visit: Date of the latest visit.
State: Where the patch stands on the visit rotation.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Visit: The public alias of one visit.
Patch: The patch the visit belongs to.
Date: The visit date.
Plant: The accepted plant name.
Removal band: How much of the patch the crew removed, decoded, with its recorded band.
Green cover: Share of the plant still green.
Crew hours: Hours spent on the visit.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Record: The public alias of one treatment method record.
Visit: The visit the method belongs to.
Patch: The patch treated.
Method: The treatment method, decoded.
Herbicide: Whether the method uses herbicide.
Product: The herbicide product, where one was used.
Concentration: Herbicide concentration, where recorded.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Recorded label: A plant name as entered in the field.
Accepted plant: The accepted name it resolves to.
Resolution: How the entered name resolves.
Family state: Whether the name family is still open for review.
Records: Records carrying that entered name.
State: Whether the name was observed in the records.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Table: The record the field belongs to.
Field: The form field.
Applies: When the field is on the form.
State: How a blank value in the field should be read.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Grain: The record type counted.
Field: The field counted.
State: Recorded, not recorded or not applicable.
Records: Records in that state.
Denominator: Records the field applies to.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

What the columns mean

Domain: The coded list a field draws from.
Code: The stored code.
Label: What the code means.
Active: Whether the code is still offered on the form.
Data status: Demo data on every row. In the file; left off the screen.
Record class: GENERATED_SAMPLE on every row. In the file; left off the screen.

Invasive Plants

Reports

One question per report, each checked against its own totals.

Report

Invasive Plants

Methods

Where a screen and this page differ, one of them has a defect.

How every figure on these screens is computed

Every record here is generated and runs through 17 Aug 2026. Each section below names the screen it explains; every number in it is read from the same setting or count the screen uses.

Action center

The return-visit list

What is drawn: One row per mapped patch that is due a visit: 246 of 520 mapped patches.
When it appears: A patch is listed under the first due state that applies, in this order: cleared then past its recheck, past the rotation ceiling, never visited. Each patch appears once.
What counts as overdue: A patch whose last record was a full clearance is rechecked within 365 days; any other patch is due after 1,095 days with no visit. Days are counted to 17 Aug 2026.
What would throw it off: A visit filed against the wrong patch, or a full clearance recorded as partial, puts a patch on the wrong clock. A long gap means nobody has looked; it is not evidence of plants, or of their absence.

Outcomes

Clearance follow-up

What is drawn: For each record of a full clearance, the next dated visit to the same patch and what it recorded.
When it appears: Only where a later dated visit exists. A clearance with no later visit is shown as not yet revisited, never as a success.
What counts as still clear: The cover class domain runs from a trace under one percent up to high, over fifty percent, and its floor is a trace rather than a zero: the form carries no code meaning the plant is gone, and the only recordable stand-in for still-clear is a removal band of none removed, which cannot separate nothing was there from nobody treated it this time.
What would throw it off: A later visit filed against a neighboring patch reads as a regrowth, or a clean return, that never happened.

Map

Map marks

What is drawn: Each patch at its recorded position: 331 mapped patches draw their boundary at close zoom with a center square, and 189 point records draw as clustered points.
When it appears: A patch whose boundary is unavailable draws a hollow dashed center marker instead of a boundary.
What counts as an area: Only a mapped boundary carries an area. A point record has none, so a hectare total covers 331 of 520 patches and says so.
What would throw it off: A boundary digitized in the wrong place moves the patch and not its visits. The map draws each record as filed.

Data & methods: Reports

Activity by year

What is drawn: Visits, patches, crew hours and treated area per calendar year, from 1,235 visit records.
When it appears: A visit counts in the year of its recorded date. A visit with no readable date is left out and counted in the report's own check.
What counts as treated area: The figure written on the visit, never the patch's mapped area. A patch nobody visited has no treated area at all, which is different from zero.
What would throw it off: Crew hours and treated area are summed as recorded, so one mistyped entry moves a whole year's total.

Data & methods: Reports

Reports

What is drawn: Each report prints one table and, above it, a second computation of its own totals.
When it appears: Every time the pane opens, from the records as loaded (through 17 Aug 2026); a report that cannot be prepared says so instead of printing zeros.
What counts as agreement: Two totals agree when they are equal, or when a part does not exceed its whole. Every pair is listed with both values under The pairs compared.
What would throw it off: A record that belongs to no patch breaks a pair, and the report then leads with the disagreement rather than the table.

What these records can and cannot support

Read from the claim policy every figure on these screens obeys: 12 of 23 quantities can be drawn for at least one record group, 11 are refused outright, and 9 are refused for some groups.

These records support (12)

Counts of visit records Every record group.
Counts of mapped patches Every record group.
Recorded cover class Record groups: mapped area, point only, no patch.
Recorded removal band Record groups: mapped area, point only, no patch.
Plant-name reconciliation Every record group.
Days between visits to one patch Record groups: mapped area, point only.
Treated area written on a visit Record groups: mapped area, point only, no patch.
A total of mapped area Record groups: mapped area, never visited.
The treated share of a patch Record groups: mapped area. Caveat for mapped area: A treated share is drawn only where the patch itself carries a recorded area, and a handful of records report treating more ground than their own patch is recorded as covering.
What the next visit recorded after a full clearance Record groups: mapped area, point only. Caveat for mapped area, point only: The cover class domain runs from a trace under one percent up to high, over fifty percent, and its floor is a trace rather than a zero: the form carries no code meaning the plant is gone, and the only recordable stand-in for still-clear is a removal band of none removed, which cannot separate nothing was there from nobody treated it this time.
Months of the year a plant was worked Record groups: mapped area, point only, no patch. Caveat for mapped area, point only, no patch: This is when work was recorded, not when the plants grow, and four months of the year hold nothing at all in most of the years these records cover.
Disagreement between two records of one patch Record groups: mapped area, point only, no patch. Caveat for mapped area, point only, no patch: The paired records are the occasions somebody happened to schedule twice rather than a random sample of visits, and there are few enough of them that no rate taken off them generalises.

These records do not support (11)

A treatment effect Mapped area: Treatment effect is not computed. The generated corpus's association check found no evidence that later recorded cover differed by removal band, but a large p-value does not establish equivalence and these observational records cannot identify an effect. Point only: Treatment effect is not computed, for the reason above. No patch: Treatment effect is not computed. There is no next visit on the same patch to read an outcome from. Never visited: Treatment effect is not computed. There is no treatment.
Retreatment efficacy Mapped area: Re-treatment efficacy is not computed. This generated corpus showed no evidence that later recorded cover was associated with the gap bucket, but that is not an equivalence test and cannot support a causal shorter-is-better reading. Point only: Re-treatment efficacy is not computed, for the reason above. No patch: No patch, so no gap and no outcome. Never visited: No visit, so no gap and no outcome.
Method efficacy Method efficacy is not computed on any arm. The method is confounded by indication: crews picked it from the patch in front of them.
A cover trend over time Mapped area: No patch cover trajectory is drawn. Cover class carries no patch identity in these records, so a line joining one patch's readings would draw a story into independent draws. Point only: No patch cover trajectory is drawn, for the reason above. No patch: No patch, so no per-patch history. Never visited: No visit, so no readings to join.
Ground treated per crew-hour Mapped area: No hectares-per-crew-hour figure is computed. Crew hours carry essentially no relationship to treated area here, so a productivity table would rank crews on noise. Point only and no patch: No hectares-per-crew-hour figure is computed, for the reason above. Never visited: No visit, so no crew hours.
Which patch replaced which Mapped area: No lineage arrow, connector or redrawn-polygon story is drawn. A successor here is commonly a different species, in a different unit, mapped earlier than the patch it supersedes, so drawing the link would invent the relationship rather than show it. Point only and never visited: No lineage is drawn, for the reason above. No patch: No patch, so no lineage.
A response to weather Mapped area: No rainfall or temperature axis is drawn. The join exists and was looked at; recorded green carries no relationship to antecedent precipitation or to maximum temperature in these records, and a covariate axis over that would be a picture of nothing. Point only and no patch: No weather axis is drawn, for the reason above. Never visited: No visit, so no recorded green to relate to weather.
A terrain effect Mapped area: No slope or aspect reading is offered. Aspect is indistinguishable from uniform and the slope medians sit within a few points of each other across every species, so a terrain contrast would be a contrast between rounding errors. Point only and never visited: No terrain reading is offered, for the reason above. No patch: No patch, so no slope and no aspect.
Herbicide results by plant Mapped area: No herbicide-by-species breakdown is drawn. The recorded active ingredients do not track the plant they were applied to, so the table would read as a prescribing pattern the records do not contain. Point only and no patch: No herbicide-by-species breakdown is drawn, for the reason above. Never visited: No visit, so no application.
A data-check flag rate Mapped area: No flag rate is computed anywhere in this program. The flags are quoted as corroboration beside a figure and are never divided by anything, because the count they would be divided by is not the count they were drawn from. Point only and no patch and never visited: No flag rate is computed, for the reason above.
A mapped-then-worked timeline Mapped area: No lifecycle or time-since-mapping reading is offered. The map date and the first visit are not ordered in these records, so an axis running from one to the other would run backwards for a great many patches. Point only: No lifecycle reading is offered, for the reason above. No patch: No patch, so no map date. Never visited: No lifecycle reading is offered: the patch has a map date and no fieldwork to place against it.

Refused for some record groups only (9)

Recorded cover class Refused for: never visited. A patch with no visit recorded no cover class, because nobody was there to record one. That is the absence of a record, never a cover of zero.
Recorded removal band Refused for: never visited. A patch with no visit recorded no removal band. Nothing was removed and nothing was left standing; there is no record either way.
Days between visits to one patch Refused for: no patch, never visited. No patch: No patch means no next visit on the same patch. There is nothing to measure a gap against, and pairing two unlinked visits by date would invent a patch history the records do not carry. Never visited: No visit means no interval. A patch nobody visited has not waited a long time between visits; it has no visits at all, and those are different facts.
Treated area written on a visit Refused for: never visited. No visit, so no treatment, so no treated area. Not zero hectares treated: no record of treatment at all.
A total of mapped area Refused for: point only, no patch. Point only: A point has no area. Not an unknown area, not a small area, not zero: a point record was never a measurement of ground. Imputing one, or counting it as zero, reports a hectare total over a denominator that has quietly dropped a third of the records. No patch: No patch, so no mapped area to total.
The treated share of a patch Refused for: point only, no patch, never visited. Point only: The denominator does not exist. A point record carries no area, so there is no share of it to take. No patch: No patch, so no patch area, so no fraction. Never visited: No visit, so no treated area, so no fraction.
What the next visit recorded after a full clearance Refused for: no patch, never visited. No patch: No patch, so there is no later visit to the same patch and no outcome to read. Never visited: No visit, so nothing was ever recorded cleared here.
Months of the year a plant was worked Refused for: never visited. No visit, so no month of work.
Disagreement between two records of one patch Refused for: never visited. No visit, so no second opinion about it.

Questions this data can answer

Choose one question. Each answer shows what was counted, what was left out, and the decision it can support.

All questions

Choose a question