Skip to content

Commit

Permalink
Merge pull request #347 from influxdata/fix-derivative
Browse files Browse the repository at this point in the history 
Correct DERIVATIVE() and GROUP BY doc
  • Loading branch information
Regan Kuchan committed Mar 29, 2016
2 parents 4255f52 + fd9659b commit ab2822b
Show file tree
Hide file tree
Showing 2 changed files with 27 additions and 10 deletions.
18 changes: 13 additions & 5 deletions content/influxdb/v0.10/query_language/functions.md
View file
Expand Up @@ -949,7 +949,11 @@ Valid time specifications for `unit` are:
`w` weeks

`DERIVATIVE()` also works with a nested function coupled with a `GROUP BY time()` clause.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause and then carries out the same procedure outlined above.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause .
It then calculates the difference between chronological field values and converts
those results into the rate of change per `unit`.
The `unit` argument is optional and, if not specified, defaults to the same
interval as the `GROUP BY time()` interval.

The `DERIVATIVE()` query with an aggregation function and `GROUP BY time()` clause:
```sql
Expand Down Expand Up @@ -1066,11 +1070,11 @@ This returns the rate of change per 12 minutes from `2015-08-18T00:00:00Z` to `2
> **Note:** Specifying `12m` as the `unit` **does not** mean that InfluxDB calculates the rate of change for every 12 minute interval of data.
Instead, InfluxDB calculates the rate of change per 12 minutes for each interval of valid data.

* `DERIVATIVE()` with two arguments, a function, and a `GROUP BY time()` clause:
* `DERIVATIVE()` with one argument, a function, and a `GROUP BY time()` clause:
Select the `MAX()` value at 12 minute intervals and calculate the rate of change per 12 minutes

```sql
> SELECT DERIVATIVE(MAX(water_level),12m) FROM h2o_feet WHERE location = 'santa_monica' AND time >= '2015-08-18T00:00:00Z' AND time < '2015-08-18T00:36:00Z' GROUP BY time(12m)
> SELECT DERIVATIVE(MAX(water_level)) FROM h2o_feet WHERE location = 'santa_monica' AND time >= '2015-08-18T00:00:00Z' AND time < '2015-08-18T00:36:00Z' GROUP BY time(12m)
```

CLI response:
Expand All @@ -1093,7 +1097,7 @@ time max
2015-08-18T00:24:00Z 2.051
```

Second, InfluxDB calculates the rate of change per `unit` (`12m`) to get the results in the `derivative` column above.
Second, InfluxDB calculates the rate of change per `12m` (the same interval as the `GROUP BY time()` interval) to get the results in the `derivative` column above.
The calculation of the first value in the `derivative` column looks like this:
<br>
<br>
Expand Down Expand Up @@ -1179,7 +1183,11 @@ Valid time specifications for `unit` are:
`w` weeks

`NON_NEGATIVE_DERIVATIVE()` also works with a nested function coupled with a `GROUP BY time()` clause.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause and then carries out the same procedure outlined above.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause.
It then calculates the difference between chronological field values and
converts those results into the rate of change per `unit`.
The `unit` argument is optional and, if not specified, defaults to the same
interval as the `GROUP BY time()` interval.

The `NON_NEGATIVE_DERIVATIVE()` query with an aggregation function and `GROUP BY time()` clause:
```sql
Expand Down
19 changes: 14 additions & 5 deletions content/influxdb/v0.11/query_language/functions.md
View file
Expand Up @@ -954,7 +954,12 @@ Valid time specifications for `unit` are:
`w` weeks

`DERIVATIVE()` also works with a nested function coupled with a `GROUP BY time()` clause.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause and then carries out the same procedure outlined above.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause.
It then calculates the difference between chronological field values and
converts those results into the rate of change per `unit`.
The `unit` argument is optional and, if not specified, defaults to the same
interval as the `GROUP BY time()` interval.


The `DERIVATIVE()` query with an aggregation function and `GROUP BY time()` clause:
```sql
Expand Down Expand Up @@ -1071,11 +1076,11 @@ This returns the rate of change per 12 minutes from `2015-08-18T00:00:00Z` to `2
> **Note:** Specifying `12m` as the `unit` **does not** mean that InfluxDB calculates the rate of change for every 12 minute interval of data.
Instead, InfluxDB calculates the rate of change per 12 minutes for each interval of valid data.

* `DERIVATIVE()` with two arguments, a function, and a `GROUP BY time()` clause:
* `DERIVATIVE()` with one argument, a function, and a `GROUP BY time()` clause:
Select the `MAX()` value at 12 minute intervals and calculate the rate of change per 12 minutes

```sql
> SELECT DERIVATIVE(MAX(water_level),12m) FROM h2o_feet WHERE location = 'santa_monica' AND time >= '2015-08-18T00:00:00Z' AND time < '2015-08-18T00:36:00Z' GROUP BY time(12m)
> SELECT DERIVATIVE(MAX(water_level)) FROM h2o_feet WHERE location = 'santa_monica' AND time >= '2015-08-18T00:00:00Z' AND time < '2015-08-18T00:36:00Z' GROUP BY time(12m)
```

CLI response:
Expand All @@ -1098,7 +1103,7 @@ time max
2015-08-18T00:24:00Z 2.051
```

Second, InfluxDB calculates the rate of change per `unit` (`12m`) to get the results in the `derivative` column above.
Second, InfluxDB calculates the rate of change per `12m` (the same interval as the `GROUP BY time()` interval) to get the results in the `derivative` column above.
The calculation of the first value in the `derivative` column looks like this:
<br>
<br>
Expand Down Expand Up @@ -1184,7 +1189,11 @@ Valid time specifications for `unit` are:
`w` weeks

`NON_NEGATIVE_DERIVATIVE()` also works with a nested function coupled with a `GROUP BY time()` clause.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause and then carries out the same procedure outlined above.
For queries that include those options, InfluxDB first performs the aggregation, selection, or transformation across the time interval specified in the `GROUP BY time()` clause.
It then calculates the difference between chronological field values and
converts those results into the rate of change per `unit`.
The `unit` argument is optional and, if not specified, defaults to the same
interval as the `GROUP BY time()` interval.

The `NON_NEGATIVE_DERIVATIVE()` query with an aggregation function and `GROUP BY time()` clause:
```sql
Expand Down

0 comments on commit ab2822b

Please sign in to comment.