MATLAB Central Newsreader - Timeseries function duplicate records error

Timeseries function duplicate records error

Bastian — Thu, 29 Oct 2009 18:51:03 -0400

When I try to create a new timeseries object using the following code (read frmo a text file):

    DateTime = FileContent{1,1}; %example: 07/30/07 15:32:35.0
    Temp = FileContent{1,2}; %example: 24.605
    ts = timeseries(Temp,DateTime);

I get the following warning and error:

   Warning: Duplicate records are detected and have been merged.
   ??? Error using ==> timeseries.init at 312
   Mismatch between the size of the data and the quality properties.
   Error in ==> timeseries.timeseries>timeseries.timeseries at 216
                  this = init(this,varargin{:});
   Error in ==> ThermoSTATv2_Oct29_2009>pbtReadFile_Callback at 125
    ts = timeseries(Temp,DateTime);

It looks like MATLAB has detected a duplicate record and has "fixed" the problem by "merging" the records. Then something happens that some other part of the timeseries function does not like because of the previous "fix". Does anybody have any clever ideas on how to prevent this from happening? or some sort of work around?

Re: Timeseries function duplicate records error

NZTideMan — Thu, 29 Oct 2009 20:11:26 -0400

On Oct 30, 7:51 am, "Bastian "
<bastianschm...@trentu.removethistext.ca> wrote:
> When I try to create a new timeseries object using the following code (read frmo a text file):
>
> DateTime = FileContent{1,1}; %example: 07/30/07 15:32:35.0
> Temp = FileContent{1,2}; %example: 24.605
> ts = timeseries(Temp,DateTime);
>
> I get the following warning and error:
>
> Warning: Duplicate records are detected and have been merged.
> ??? Error using ==> timeseries.init at 312
> Mismatch between the size of the data and the quality properties.
> Error in ==> timeseries.timeseries>timeseries.timeseries at 216
> this = init(this,varargin{:});
> Error in ==> ThermoSTATv2_Oct29_2009>pbtReadFile_Callback at 125
> ts = timeseries(Temp,DateTime);
>
> It looks like MATLAB has detected a duplicate record and has "fixed" the problem by "merging" the records. Then something happens that some other part of the timeseries function does not like because of the previous "fix". Does anybody have any clever ideas on how to prevent this from happening? or some sort of work around?

Before going into timeseries, use the function unique on DateTime to
remove duplications of time. You will also need to eliminate the
corresponding rows in Temp:
[DateTimeNew,indx]=unique(DateTime);
TempNew=Temp(indx);

Re: Timeseries function duplicate records error

Bastian — Thu, 29 Oct 2009 20:29:03 -0400

NZTideMan <mulgor@gmail.com> wrote in message <5b388161-e844-4448-ae52-b6ce29dab679@y32g2000prd.googlegroups.com>...
> On Oct 30, 7:51?am, "Bastian "
> <bastianschm...@trentu.removethistext.ca> wrote:
> > When I try to create a new timeseries object using the following code (read frmo a text file): ?
> >
> > ? ? DateTime = FileContent{1,1}; %example: 07/30/07 15:32:35.0
> > ? ? Temp = FileContent{1,2}; %example: 24.605
> > ? ? ts = timeseries(Temp,DateTime);
> >
> > I get the following warning and error:
> >
> > ? ?Warning: Duplicate records are detected and have been merged.
> > ? ???? Error using ==> timeseries.init at 312
> > ? ?Mismatch between the size of the data and the quality properties.
> > ? ?Error in ==> timeseries.timeseries>timeseries.timeseries at 216
> > ? ? ? ? ? ? ? ? ? this = init(this,varargin{:});
> > ? ?Error in ==> ThermoSTATv2_Oct29_2009>pbtReadFile_Callback at 125
> > ? ? ts = timeseries(Temp,DateTime);
> >
> > It looks like MATLAB has detected a duplicate record and has "fixed" the problem by "merging" the records. Then something happens that some other part of the timeseries function does not like because of the previous "fix". Does anybody have any clever ideas on how to prevent this from happening? or some sort of work around?
>
> Before going into timeseries, use the function unique on DateTime to
> remove duplications of time. You will also need to eliminate the
> corresponding rows in Temp:
> [DateTimeNew,indx]=unique(DateTime);
> TempNew=Temp(indx);

Thanks, that's done it. I had been trying to make use of the "unique" function but was stumped by the fact that it sorts the output. I now realize that this does not matter since timeseries puts the times back in order automatically! Thanks again for your help.