Hi all,
OK, I’ll get you some data to play with. Our data is from street-level
collections of sensor readings from a moving vehicle. We do that to avoid the
costs involved in placing hundreds of thousands of sensor packages. We might
place some sensors at high-value locations later.
Please note that I tried some predictions on air quality data from the US EPA
earlier in the year without much success. It turns out that the EPA has very
few actual sensor stations, and the data they provide is highly interpolated. I
was unable to predict an air quality index (AQI) based solely upon historical
AQI readings from a particular location.
It might be possible, though, to better predict AQI if stations upwind of your
location were taken into account. I haven’t tried that yet.
If you want to play with the US EPA data, try this:
[[
$ mkdir airnow; cd airnow
$ for i in {1990..2014}; do curl
http://aqsdr1.epa.gov/aqsweb/aqstmp/airdata/daily_81102_$i.zip >
daily_81102_$i.zip; done
$ for i in `ls -1`; do unzip $i; done
$ mkdir zipfiles; mv *.zip zipfiles
]]
In the above, the magic number 81102 indicates a particular sensor, in this
case particulate matter greater in size than ten microns (PM10 Mass). Other
sensors are available. 25 years worth of data are then grabbed.
Please see this URL to get other data from the EPA’s AirNow program:
http://www.epa.gov/airquality/airdata/ad_data.html
<http://www.epa.gov/airquality/airdata/ad_data.html>
Regards,
Dave
--
http://about.me/david_wood
> On Aug 19, 2015, at 02:46, Pascal Weinberger <[email protected]>
> wrote:
>
> I would love some of that as well!
> For nostradamiq.org <http://nostradamiq.org/> , Could it be possible to get
> some realtimeish data as well? :))))
>
> Best,
>
> Pascal
>
> On Aug 19, 2015 4:56 AM, "Matthew Taylor" <[email protected]
> <mailto:[email protected]>> wrote:
> Yes, that would be great!
>
> ---------
> Matt Taylor
> OS Community Flag-Bearer
> Numenta
>
> On Tue, Aug 18, 2015 at 7:29 PM, David Wood <[email protected]
> <mailto:[email protected]>> wrote:
> Hey, we have a bunch of temporal data like this: WE are collecting air
> quality data from a variety of sensors. Each reading is geo- and
> temporally-tagged.
>
> I can put some of this on the public Web if people want to play. Let me know.
>
> Regards,
> Dave
> --
> http://about.me/david_wood <http://about.me/david_wood>
>
>
>
>> On Aug 18, 2015, at 18:38, Matthew Taylor <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>> Hello NuPIC,
>>
>> Jeff Fohl just pointed me to this: https://data.sparkfun.com/streams/
>> <https://data.sparkfun.com/streams/>
>>
>> Apparently, this is like River View for IoT! Take a look at all the temporal
>> data they have, it's like we're kids in a giant instrumented candy store.
>>
>> Regards,
>>
>> ---------
>> Matt Taylor
>> OS Community Flag-Bearer
>> Numenta
>
>
