XMS - time of begin and end return strange number

[Lukas_Doubek]

Hi, im calling the XMS function on channel

The first option "Calculation within one peak" returns times in ms, as the input channel is waveform with waveform unit setted to ms.

All good.

But when i change it to "Calculation including all peaks", it return 1000x bigger number. Its normal behaviour?

check the screens 🙂

[Brad_Turpin]

Hey Lukas,

What version of DIAdem are you using?  Do you see this effect with all data files you've tried or just certain ones?

Brad Turpin

DIAdem Product Support Engineer

National Instruments

[Lukas_Doubek]

Hi, 2017 64bit. And im not alolen, my colleagues have the same behaviour, when i noticed them, they told me they know about it and they are used to it instead of report it to me 🙂

[Brad_Turpin]

Hey Lukas,

I mentioned this to R&D, and they have asked if you would submit a data set that shows this effect, either to this post or to brad.turpin@ni.com.

Then they'll investigate,

Brad Turpin

DIAdem Product Support Engineer

National Instruments

[Lukas_Doubek]

Hi, 

Try running this command

Call ChnXMSCalc("", "[1]/11HEAD0000WSACRP" , 3, 0, "in multiple peaks", false)

 The result if 34.8474113155008 which i recognised as true result.

Then try convert the waveform x-unit to "ms".

after runnig the same command, i got the result: 39.2131886153666 which i think is false. imho it should return the same result, because the input parameter is specified like this "

Specifies the time X in milliseconds

". (((it should be written smtin like this "if you have channel in ms, specifies the X in ms*1000, other case just ms*1" )))" i think the program should deal with it, or have it written in help, that the input channel must meet the specs to have the waveform in specified units. but now its not specified and i the different results got me confused, cause i dont know what to say to my colleagues who use my program that we have different results. 

i tested it now in Diadem 2019 but there is the same behaviour..

We will soon start to use Diadem 2019 so it could be good to have it fixed. Thanks for info how it goes.

[Lukas_Doubek]

i added attachement, tdm file in zip archive, but i dont see it, tell me if you see it and download it or i will reupload it

[Lukas_Doubek]

I got another strange thing with this function.

When i add novalues to numeric channel on which i want to count Xms function, it doesnt create result, no properties, even not the tree of properties starting with "Result". Nothing there.

It tells me no message that its not going to count it. I spent time searching for the mistake i make, but i think its a bug of diadem.

Sometimes channels have error peaks, and we cleart them by putting NOVALUE values on that interval, but your function doesnt work after. 

Can you describe me whats the problem? I need it to work on channels whjich have NOVALUEs as well.

Should i create its own topic for this problem?

// When i interpolate the interval, it works as there is no NOVALUE value, but thats not the case i need.

[Walter_Rick]

Hi Lukas,

I have tested both behaviors (high ms values and Novalue) in DIAdem 2017 SP1 (64 bit) and DIAdem 2019 SP1 with the dataset you posted (forBrad.zip, one channel with 3001 values). In both cases the results are OK.

Greetings

Walter

[Lukas_Doubek]

Nope, you mixed 2 things together.

I reported 2 erros:

1) for channels with waveform units with "ms", the function returns false time intervals as shown on the picture.

In  the zip file, there is one channel:

When i run XMS function on this channel, the resuts are fine:

Ok now lets convert the waveform unit from "s" to "ms". Thats the point wich causes the behaviour. Change it like i did:

And when i run the same function, now the time resutls are nonsense. It should be just 50ms and not 50000ms. Thats the problem i reported at the first, dont mix it with the NoValues problem.

I will explain the NoValues problem in separate thread, i will make new thread as it has nothing commin with this problem.

[Lukas_Doubek]

Send me please screen of the channel values of the channel HEA_NV for me to see that you really have NOVALUE values inside. You posted just screen of the channel unit with "g" but i cant see that you have novalues there. EWven the preview of the channel i dont see empty places on the curve.

On this picture i posted, you can see that i have NOVALUE value in the channel which cause me the problem.