• State Verified Answer
  • Replies 15 replies
  • Subscribers 16 subscribers
  • Views 1864 views
  • Users 0 members are here
Options
Related

Toybox.SensorHistory.getPressureHistory({:period => 10, :order => 1}); does not compile

Drake1969
over 2 years ago
Toybox.SensorHistory.getPressureHistory({}); this is going to be compiled the one with options not
  • +1 over 2 years ago

    What's the error you see? If it's from type checking, use ORDER_NEWEST_FIRST or ORDER_OLDEST_FIRST and not a Number for order

  • 0 over 2 years ago in reply to jim_m_58

    ops.. you are right, now the compile error is gone. Is there any way to know the number of elements of the history?

  • +1 over 2 years ago in reply to Drake1969

    You can walk the iterator and count them.  You may find you have two cases where you use the info from SensorHistory.get*History - you either want only the newest sample, or you want to graph what's there.

    In the case of the graph, you'll know the max width of the graph.  In that case you just walk the iterator until you hit the max width of the graph or you run out of samples.

  • 0 over 2 years ago in reply to jim_m_58

    i tried to count items inside an HR history of the fenix 5. I debugged and found that records are 120. Unfortunately when i'm counting if i retrieve the value... i hit the:

    Error: Unexpected Type Error Details: Failed invoking <symbol> 

    because the next() went over the 120th record...

    just for debug i tried the following and the 120 record invoke the error...

                var j=0;
                    while (true)
                    {
                        System.println(j);
                        value =((HRHist.next().data));   <----- error is here
                        if ((value==null) || (value==ActivityMonitor.INVALID_HR_SAMPLE)) {value =0;}  
                        TempVect[j]=value;
                      j++;
                      if (j>=HistorySamples)
                      {break;}
                    }    
  • 0 over 2 years ago in reply to Drake1969

    This is something where the sim is different than real devices.  The number of samples differ as does the time between samples.  What happens if you see less than 120 sample in the sim?  You don't want to use a count - you want to go until "next()" is null.  When you first turn on a watch, you might only have one sample

  • 0 over 2 years ago in reply to jim_m_58

    maybe you didnt get the point:

    when i execute 

    value =((HRHist.next().data));   <----- error is here

    the "value" didnt get the null ... the app just crash there... 

    Well i can say that the fenix 5 has 180 samples for temp, press and elev but only 120 for Heart Rate... 

  • 0 over 2 years ago in reply to Drake1969

    No you have a bug here:

    value =((HRHist.next().data));

    HRHist.next() will be nulll when you reach the end and doing "null.data" throws an error.

    You want to null check HRHist().next() and only if it's not null do you want to use .data

    var a =HRHist.next();

    if(a!=null) {value=a.data;}

    A clearer way to do this is:

    		//iter is what you get back from the get*History() call
		if(iter!=null) {
			var sample=iter.next();
            var count=0;
			while(sample!=null && count<100) {   //you can add a counter and use at most, say 100 sampples here
				var data=sample.data;
				if(data!=null) {
                    //do what you want with the data
				}
				count++;
				sample=iter.next();
			}
		}

  • 0 over 2 years ago in reply to jim_m_58

    i apologize... i didnt get the point!!! ok.. going to work on it to change my code

  • +1 over 2 years ago in reply to Drake1969
    Drake1969 said:
    Well i can say that the fenix 5 has 180 samples for temp, press and elev but only 120 for Heart Rate... 

    Turn off the f5 turn it back on, and check again.  You'll see far lower numbers.  And on different devices, it won't top out at the same numbers.  That's why you want to null check.

  • 0 over 2 years ago in reply to jim_m_58

    yes correct... 

    i did the homework, i find the history length first and then show the data into graphs. Looks like working like a charm. I must say that i didnt carefully read the documentation... your example is better than the one into doc!

    thank you very much. My watchface is almost ready... (camuflage watchface)