Re: Mathematica 5,6,7 WeatherData, CityData
- To: mathgroup at smc.vnet.net
- Subject: [mg94142] Re: Mathematica 5,6,7 WeatherData, CityData
- From: Mac <mwjdavidson at googlemail.com>
- Date: Fri, 5 Dec 2008 07:04:20 -0500 (EST)
- References: <gh5nqh$odq$1@smc.vnet.net> <gh8hgg$qt8$1@smc.vnet.net>
I'm also very impressed with the possibilities offered by the new
weather station function in Mathematica 7.0. The geographical and
environmental data functions for Mathematica are growing quickly and
valuable (at least to me).
My guess is for the city data function that Wolfram is limited by the
data source that they are using. It is not clear to me how much
Wolfram is actively contributing to filling gaps in the dataset as
coordinates for cities can be derived from a number of locations. I've
proposed data for some missing cities to them but never got feedback.
Out of interest I tried the following functions to determine what
percentage of cities in CityData have coordinates.
In[61]:= allcities =
Flatten[CityData[{All, #}] & /@ CountryData[] , {1, 2}];
In[62]:= allcities // Dimensions
Out[62]= {163391, 3}
In[63]:= DeleteCases[
CityData[#, "Coordinates"] & /@
allcities, {_Missing, _Missing}] // Dimensions
Out[63]= {148733, 2}
In short - about 90% (or 148733/163391*100) of cities have coordinates
associated with them. Deventer must belong to the 10% without
coordinates.
Still this is not too bad and certainly a lot of useful data.
Mac
On Dec 4, 1:13 pm, "sjoerd.c.devr... at gmail.com"
<sjoerd.c.devr... at gmail.com> wrote:
> You're right here.
>
> The "Shape" properties of The Netherlands falls woefully short. Ik
> looks a bit like a 1930 map or so. The "FullPolygon" data set is
> somewhat better, with the large Waddenzee islands present, though one
> of the IJsselmeer polders is still missing.
>
> I suggest you collect your own contour using a google earth bitmap and
> mathematica's coordinate picker tool. Shouldn't be too difficult, but
> getting lat/longs right might be more complex.
>
> Deventer has indeed its coordinates missing from the database.
> Mathematica does know its population though. Not very satisfactory, I
> agree.
>
> Cheers -- Sjoerd
>
> On Dec 3, 12:42 pm, P_ter <petervansumme... at gmail.com> wrote:> Hello,
>
> > I was impressed by an example for Mathematica 7.0, that is, the globe w=
it=
>
> h theweatherstations. So, I wanted to make something like that for the Ne=
=
> therlands.> First of all, have a look atwww.maps.google.nltocompare that =
map with th=
>
> e map from the CountryData: islands are missing, a famous dike is missing=
, =
> and a piece of land won from the sea is missing. So making a polygon for =
th=
> e Netherlands does not work.
>
> > How to find all theweatherstations in the Netherlands with Mathematica?=
=
>
> There is no method to determine if aweatherstation is in a certain cou=
nt=
> ry.> It is here that I stumbled also on holes in CityData: CityData[{All,=
"Ove=
>
> rijssel", "Netherlands"}] shows that the city Deventer is in it.> But whe=
n I do: CityData[{"Deventer", "Netherlands"}, "Coordinates"],
> > I get: {Missing["NotAvailable"], Missing["NotAvailable"]}
>
> > It means that the missing coordinates make it difficult to combine the =
av=
>
> ailable CityData and WeatherData as to determine if aweatherstation is (n=
=
> ear a city) in the Netherlands.
>
>
>
>
>
>
>
> > ??
> > with friendly greetings,
> > P_ter
- Follow-Ups:
- RE: Re: Mathematica 5,6,7 WeatherData, CityData
- From: "Tugrul Temel" <temelt@xs4all.nl>
- RE: Re: Mathematica 5,6,7 WeatherData, CityData