Documentation: performance_results
Limite de débit : 300 requêtes par minute. Une fois la limite dépassée, toute requête ultérieure entraînera une réponse 429 (Trop de requêtes).
| Nom de la ligne | Type | Exemple |
|---|---|---|
|
uid subject_origin + subject_uid + period_label + metric; Unique identifier of a single result. A result is updated in place when its period is scored again, until it is frozen; its earlier values are kept in the history.
|
string |
installations_ts_123_victron_VRM12345_2026-06-14_ec |
|
subject_origin The table the subject was scored from, so a uid can be followed back to what it names.
|
string |
installations_ts |
|
subject_uid The subject scored, named as the table given in subject_origin names it.
|
string |
123_victron_VRM12345 |
|
period_type The kind of period scored, e.g. day or quarter. A period is given by its type and the moment it starts, so that periods of different lengths sit in the one table.
|
enum: day, week, month, quarter, year |
day |
|
period_start The moment the period scored starts, in UTC (a period cut on a local clock starts at that clock's midnight).
|
datetime |
2026-06-13T22:00:00+00:00 |
|
period_label The period scored as a person would name it. For example, it could be built from period_start and period_type via the dataworker's build_period_label utility, on the clock the period was cut on.
|
string |
2026-Q2 |
|
is_frozen Whether the result is final. A frozen result is no longer updated when its period is scored again, so updated_at says when it was frozen.
|
boolean |
false |
|
metric The metric scored, as the program reporting it names it.
|
string |
ec |
|
value What the subject scored on the metric, in unit_of_measurement. Only period (.) decimal delimiters are supported. Do not use commas (,) — they will cause an error.
|
float |
38400.0 |
|
increments_count How many increments the value was counted from, where it is a count rather than a measurement. A period sums these alongside the value to score it, so that increments the subject reported nothing for are left out rather than counted as failures. Only period (.) decimal delimiters are supported. Do not use commas (,) — they will cause an error.
|
integer |
48 |
|
target_value The target the value was judged against, kept as it stood when the result was scored. Only period (.) decimal delimiters are supported. Do not use commas (,) — they will cause an error.
|
float |
42000.0 |
|
unit_of_measurement Unit of measurement for the metric being tracked, such as kWh, number of connections, etc.
|
string |
number |
|
target_uid The uid of the data_targets row the target came from, so a result can be traced to the target in force when it was scored, whatever happened to targets since.
|
string |
123_c782c9d4731f26d3d040d6a4a24ebd66 |
|
result Whether the value met its target (ok), nearly did (warn) or did not (fail). Null where the metric is not scored on its own, being assessed over a longer period instead.
|
enum: ok, warn, fail |
ok |
|
readings How many readings the value was taken from, which says how much of the period the subject actually reported for. Only period (.) decimal delimiters are supported. Do not use commas (,) — they will cause an error.
|
integer |
96 |
|
created_at Timestamp when row was inserted
|
datetime |
2021-05-03 6:23:55 |
|
updated_at Timestamp when row was updated
|
datetime |
2021-05-03 6:23:55 |
|
source_id References Source of data
|
integer |
1 |
|
import_id References Import of data
|
integer |
1 |
|
organization_id References Organization of data
|
integer |
1 |
|
custom Saves additional imported data out of standard structure
|
jsonb |
{ "some": "key" } |
|
data_origin Origin of the data, generic or name of the api
|
string |
generic |
|
last_import_id Id of the last import which updated the record
|
integer |
337 |
|
last_source_id Id of the last Source which updated the record
|
integer |
33 |