Thread: stationXML and responses

Started: 2011-10-20 17:50:25
Last activity: 2011-10-21 17:21:06
Topics: Web Services
Philip Crotwell
2011-10-20 17:50:25
Hi

I am looking into pulling responses via the stationXML XML web
service, and am getting a <Response> that contains only a
<StageSensitiviy> but the docs in the xschema it should have one of
<PolesZeros>, <Coefficients>, etc. From the xschema:
<xs:documentation>Choose one of these responses. There should be one
response per stage.</xs:documentation>
</xs:annotation>

An example is <Response stage="2"> in:

http://www.iris.edu/ws/station/query?net=II&sta=AAK&loc=00&cha=BH1&timewindow=1990-11-30%2C1990-11-30&level=resp

The resp file via seismoquery for this same channel has a "polezero"
blockette with zero poles and zeros. I suppose this is essentially the
same as not having one at all, but seems confusing. Should either the
docs should be changed to allow a <Response> element without a
response inside it, or should this one should include a <PolesZeros>
with no poles or zeros?

On this subject, the xml is actually valid because the xschema doesn't
enforce the "at least one of" rule. This is because some of the
elements inside of the <xs:choice> have minOccurs="0", so it is valid
to choose to put nothing inside the choice. Just seems like a bad way
to specify the schema. The schema also doesn't limit the number, so it
would be completely valid to put 27 <PolesZeros> elements inside a
<Response>, which also seems wrong, at least by seed conventions.
Perhaps the choice should have a minOccurs=1 maxOccurs=1 on it???

I know this doesn't really effect the meaning of the response, but
small ambiguities like this make writing software to use stationXML
harder.

thanks,
Philip

  • Mary Templeton
    2011-10-20 16:54:48

    On Oct 20, 2011, at 8:50 AM, Philip Crotwell wrote:

    The resp file via seismoquery for this same channel has a "polezero"
    blockette with zero poles and zeros. I suppose this is essentially the
    same as not having one at all, but seems confusing. Should either the
    docs should be changed to allow a <Response> element without a
    response inside it, or should this one should include a <PolesZeros>
    with no poles or zeros?

    A pole/zero blockette with zero poles and zeros should be equivalent
    to an infinitely flat unity response that will get multiplied by the B58 gain
    (which would be appropriate for an A/D stage). For what that's worth...

    I've also seen this stage written with a FIR coefficient blockette (instead
    of poles and zeros) with a single coefficient of 1.0.

    I realize this begs the question of how the case gets handled....

    --Mary


    thanks,
    Philip
    _______________________________________________
    webservices mailing list
    webservices<at>iris.washington.edu
    http://www.iris.washington.edu/mailman/listinfo/webservices



  • Philip Crotwell
    2011-10-21 17:21:06
    Hi

    I realized I was incorrect about the 27 PolesZeros elements.

    to specify the schema. The schema also doesn't limit the number, so it
    would be completely valid to put 27 <PolesZeros> elements inside a
    <Response>, which also seems wrong, at least by seed conventions.
    Perhaps the choice should have a minOccurs=1 maxOccurs=1 on it???

    You cannot have 27 PolesZeros elements, but you can have an unbounded
    number of FIR elements. I presume that if there are multiple FIR
    filters, then each should be contained in a separate Response element,
    and hence the maxOccurs should not be here.
    <xs:element name="FIR" minOccurs="0" maxOccurs="unbounded"
    type="StaXml:FIRType"/>

    thanks,
    Philip


    On Thu, Oct 20, 2011 at 10:50 AM, Philip Crotwell <crotwell<at>seis.sc.edu> wrote:
    Hi

    I am looking into pulling responses via the stationXML XML web
    service, and am getting a <Response> that contains only a
    <StageSensitiviy> but the docs in the xschema it should have one of
    <PolesZeros>, <Coefficients>, etc. From the xschema:
    <xs:documentation>Choose one of these responses. There should be one
    response per stage.</xs:documentation>
    </xs:annotation>

    An example is <Response stage="2"> in:

    http://www.iris.edu/ws/station/query?net=II&sta=AAK&loc=00&cha=BH1&timewindow=1990-11-30%2C1990-11-30&level=resp

    The resp file via seismoquery for this same channel has a "polezero"
    blockette with zero poles and zeros. I suppose this is essentially the
    same as not having one at all, but seems confusing. Should either the
    docs should be changed to allow a <Response> element without a
    response inside it, or should this one should include a <PolesZeros>
    with no poles or zeros?

    On this subject, the xml is actually valid because the xschema doesn't
    enforce the "at least one of" rule. This is because some of the
    elements inside of the <xs:choice> have minOccurs="0", so it is valid
    to choose to put nothing inside the choice. Just seems like a bad way
    to specify the schema. The schema also doesn't limit the number, so it
    would be completely valid to put 27 <PolesZeros> elements inside a
    <Response>, which also seems wrong, at least by seed conventions.
    Perhaps the choice should have a minOccurs=1 maxOccurs=1 on it???

    I know this doesn't really effect the meaning of the response, but
    small ambiguities like this make writing software to use stationXML
    harder.

    thanks,
    Philip



09:14:07 v.01697673