162 lines
5.5 KiB
Markdown
162 lines
5.5 KiB
Markdown
---
|
|
title: Sample data
|
|
description: >
|
|
Use sample data to familiarize yourself with time series data and InfluxDB.
|
|
InfluxData provides many sample time series datasets to use with InfluxDB
|
|
and InfluxDB Cloud.
|
|
aliases:
|
|
- /influxdb/v2.1/write-data/sample-data/
|
|
menu: influxdb_2_1_ref
|
|
weight: 7
|
|
---
|
|
|
|
Use sample data to familiarize yourself with time series data and InfluxDB.
|
|
InfluxData provides many sample time series datasets to use with InfluxDB.
|
|
You can also use the [Flux InfluxDB sample package](/{{< latest "flux" >}}/stdlib/influxdata/influxdb/sample/)
|
|
to view, download, and output sample datasets.
|
|
|
|
- [Air sensor sample data](#air-sensor-sample-data)
|
|
- [Bird migration sample data](#bird-migration-sample-data)
|
|
- [NOAA sample data](#noaa-sample-data)
|
|
- [NOAA NDBC data](#noaa-ndbc-data)
|
|
- [NOAA water sample data](#noaa-water-sample-data)
|
|
- [USGS Earthquake data](#usgs-earthquake-data)
|
|
|
|
## Air sensor sample data
|
|
|
|
{{% caption %}}
|
|
**Size**: ~600 KB • **Updated**: every 15m
|
|
{{% /caption %}}
|
|
|
|
Air sensor sample data represents an "Internet of Things" (IoT) use case by simulating
|
|
temperature, humidity, and carbon monoxide levels for multiple rooms in a building.
|
|
|
|
To download and output the air sensor sample dataset, use the
|
|
[`sample.data()` function](/{{< latest "flux" >}}/stdlib/influxdata/influxdb/sample/data/).
|
|
|
|
```js
|
|
import "influxdata/influxdb/sample"
|
|
|
|
sample.data(set: "airSensor")
|
|
```
|
|
|
|
#### Companion SQL sensor data
|
|
The air sensor sample dataset is paired with a relational SQL dataset with meta
|
|
information about sensors in each room.
|
|
These two sample datasets are used to demonstrate
|
|
[how to join time series data and relational data with Flux](/influxdb/v2.1/query-data/flux/sql/#join-sql-data-with-data-in-influxdb)
|
|
in the [Query SQL data sources](/influxdb/v2.1/query-data/flux/sql/) guide.
|
|
|
|
<a class="btn download" href="https://influx-testdata.s3.amazonaws.com/sample-sensor-info.csv" download>Download SQL air sensor data</a>
|
|
|
|
## Bird migration sample data
|
|
|
|
{{% caption %}}
|
|
**Size**: ~1.2 MB • **Updated**: N/A
|
|
{{% /caption %}}
|
|
|
|
Bird migration sample data is adapted from the
|
|
[Movebank: Animal Tracking data set](https://www.kaggle.com/pulkit8595/movebank-animal-tracking)
|
|
and represents animal migratory movements throughout 2019.
|
|
|
|
To download and output the bird migration sample dataset, use the
|
|
[`sample.data()` function](/{{< latest "flux" >}}/stdlib/influxdata/influxdb/sample/data/).
|
|
|
|
```js
|
|
import "influxdata/influxdb/sample"
|
|
|
|
sample.data(set: "birdMigration")
|
|
```
|
|
|
|
The bird migration sample dataset is used in the [Work with geo-temporal data](/influxdb/v2.1/query-data/flux/geo/)
|
|
guide to demonstrate how to query and analyze geo-temporal data.
|
|
|
|
## NOAA sample data
|
|
|
|
There are two National Oceanic and Atmospheric Administration (NOAA) datasets
|
|
available to use with InfluxDB.
|
|
|
|
- [NOAA NDBC data](#noaa-ndbc-data)
|
|
- [NOAA water sample data](#noaa-water-sample-data)
|
|
|
|
### NOAA NDBC data
|
|
|
|
{{% caption %}}
|
|
**Size**: ~1.3 MB • **Updated**: every 15m
|
|
{{% /caption %}}
|
|
|
|
The **NOAA National Data Buoy Center (NDBC)** dataset provides the latest
|
|
observations from the NOAA NDBC network of buoys throughout the world.
|
|
Observations are updated approximately every 15 minutes.
|
|
|
|
To download and output the most recent NOAA NDBC observations, use the
|
|
[`sample.data()` function](/{{< latest "flux" >}}/stdlib/influxdata/influxdb/sample/data/).
|
|
|
|
```js
|
|
import "influxdata/influxdb/sample"
|
|
|
|
sample.data(set: "noaa")
|
|
```
|
|
|
|
{{% note %}}
|
|
#### Store historical NOAA NDBC data
|
|
|
|
The **NOAA NDBC sample dataset** only returns the most recent observations;
|
|
not historical observations.
|
|
To regularly query and store NOAA NDBC observations, add the following as an
|
|
[InfluxDB task](/influxdb/v2.1/process-data/manage-tasks/).
|
|
Replace `example-org` and `example-bucket` with your organization name and the
|
|
name of the bucket to store data in.
|
|
|
|
{{% get-shared-text "flux/noaa-ndbc-sample-task.md" %}}
|
|
{{% /note %}}
|
|
|
|
### NOAA water sample data
|
|
|
|
{{% caption %}}
|
|
**Size**: ~10 MB • **Updated**: N/A
|
|
{{% /caption %}}
|
|
|
|
The **NOAA water sample dataset** is static dataset extracted from
|
|
[NOAA Center for Operational Oceanographic Products and Services](http://tidesandcurrents.noaa.gov/stations.html) data.
|
|
The sample dataset includes 15,258 observations of water levels (ft) collected every six minutes at two stations
|
|
(Santa Monica, CA (ID 9410840) and Coyote Creek, CA (ID 9414575)) over the period
|
|
from **August 18, 2015** through **September 18, 2015**.
|
|
|
|
{{% note %}}
|
|
#### Store NOAA water sample data to avoid bandwidth usage
|
|
To avoid having to re-download this 10MB dataset every time you run a query,
|
|
we recommend that you [create a new bucket](/influxdb/v2.1/organizations/buckets/create-bucket/)
|
|
(`noaa`) and write the NOAA sample water data to it.
|
|
|
|
```js
|
|
import "experimental/csv"
|
|
|
|
csv.from(url: "https://influx-testdata.s3.amazonaws.com/noaa.csv")
|
|
|> to(bucket: "noaa", org: "example-org")
|
|
```
|
|
{{% /note %}}
|
|
|
|
The NOAA water sample dataset is used to demonstrate Flux queries in the
|
|
[Common queries](/influxdb/v2.1/query-data/common-queries/) and
|
|
[Common tasks](/influxdb/v2.1/process-data/common-tasks/) guides.
|
|
|
|
## USGS Earthquake data
|
|
|
|
{{% caption %}}
|
|
**Size**: ~6 MB • **Updated**: every 15m
|
|
{{% /caption %}}
|
|
|
|
The United States Geological Survey (USGS) earthquake dataset contains observations
|
|
collected from USGS seismic sensors around the world over the last week.
|
|
Data is updated approximately every 15m.
|
|
|
|
To download and output the last week of USGS seismic data, use the
|
|
[`sample.data()` function](/{{< latest "flux" >}}/stdlib/influxdata/influxdb/sample/data/).
|
|
|
|
```js
|
|
import "influxdata/influxdb/sample"
|
|
|
|
sample.data(set: "usgs")
|
|
```
|