Methods to describe a data set

[antz0r]

Hello,

This might be a simple question but I am trying to find a good method of describing a data set. I have a data set of voltage vs time for a cycle, where voltage varies a lot. Would taking the average of voltage over the entire number of times be a good way of describing everything? Or are there other more accurate methods of getting a single and accurate voltage number for the whole cycle?

Thanks

 

[symbolipoint]

Would tell about this more precisely:

where voltage varies a lot.

?

 

[antz0r]

The voltage indicates the thrust measurement at that time. So if I convert it to thrust, it would be around 50mlbf most of the times. However, because of thermal effects (ie the instruments heat up), the measurements slowly increase over time and decrease once the device is turned off. The problem is that I don't know the exact amount it drifts because there's really no way to measure it.

I'm trying to get a single thrust measurement number for one run cycle that can describe the data. Due to the thermal drift, it's hard to represent the measurement if I just take the average of all the instantaneous thrust measurements over time. Are there any other methods of describing the data set as accurately as possible?

 

[symbolipoint]

The voltage indicates the thrust measurement at that time. So if I convert it to thrust, it would be around 50mlbf most of the times. However, because of thermal effects (ie the instruments heat up), the measurements slowly increase over time and decrease once the device is turned off. The problem is that I don't know the exact amount it drifts because there's really no way to measure it.

I'm trying to get a single thrust measurement number for one run cycle that can describe the data. Due to the thermal drift, it's hard to represent the measurement if I just take the average of all the instantaneous thrust measurements over time. Are there any other methods of describing the data set as accurately as possible?

You seem to need MORE data. One single measurement will not give you much to use for graphing or other numeric processing. An engineer-person should probably give you well directed advice. Do you maybe have some way to create a model to predict what value should be found in measurement?