Combining Force and Velocity Data in Excel with Irregular Sample Rate

  • Thread starter Thread starter 24forChromium
  • Start date Start date
  • Tags Tags
    Excel
Click For Summary
The discussion centers on combining two data series—force over time and velocity over time—where the force data is consistently sampled every 0.1 seconds, while the velocity data is irregularly sampled due to the nature of the measurement process. The goal is to align these datasets without manually matching approximately 2000 entries. Suggestions include creating a unified time series at the smallest increment and using formulas to fit both datasets into this timeline, though this may result in a sparse dataset. A Kalman filter is proposed as an effective method for integrating the sporadic velocity measurements with the continuous force data, as it can handle irregular updates and measurement noise. Additionally, the use of the VLOOKUP function in spreadsheet software is recommended for matching values based on time, with the inclusion of the IFERROR function for calculations involving the data.
24forChromium
Messages
155
Reaction score
7
Not really a programming question, but if you will:

I have two data series, one is force over time, the other is velocity over time.

The first series has a sample rate of 1 measurement per 0.1 second and is consistent.
The second series has an irregular sample rate (due to the intermittent nature of the photogate) and the time of each measured velocity is precise to the forth decimal, although I can convert them into one decimal place if I have to, the precision there is not terribly important.

I would like to combine the dependent (force and velocity) over just one time variable, how can I do that without mannually dragging the velocity measurements to the matching place for about 2000 times?
 
Technology news on Phys.org
24forChromium said:
Not really a programming question, but if you will:

I have two data series, one is force over time, the other is velocity over time.

The first series has a sample rate of 1 measurement per 0.1 second and is consistent.
The second series has an irregular sample rate (due to the intermittent nature of the photogate) and the time of each measured velocity is precise to the forth decimal, although I can convert them into one decimal place if I have to, the precision there is not terribly important.

I would like to combine the dependent (force and velocity) over just one time variable, how can I do that without mannually dragging the velocity measurements to the matching place for about 2000 times?

I've wondered about how to best do that as well in the past. I used Google to do a search for ways to combine datasets with different time values into one spreadsheet, and this hit seems to have a pretty good answer (last answer in the thread):

http://superuser.com/questions/620810/match-multiple-time-series-columns

You make a single set of time values at your smallest increment, and then use that formula (or similar) to fit each of your two datasets into the overall time line.

The disadvantage of this method is that you will end up with a pretty sparse spreadsheet if you use the 0.1ms time steps that you mentioned. Will your data lose much information if you round up that data to the nearest 10ms, for example?
 
24forChromium said:
Not really a programming question, but if you will:

I have two data series, one is force over time, the other is velocity over time.

The first series has a sample rate of 1 measurement per 0.1 second and is consistent.
The second series has an irregular sample rate (due to the intermittent nature of the photogate) and the time of each measured velocity is precise to the forth decimal, although I can convert them into one decimal place if I have to, the precision there is not terribly important.

I would like to combine the dependent (force and velocity) over just one time variable, how can I do that without mannually dragging the velocity measurements to the matching place for about 2000 times?
An estimation technique such as a Kalman filter would work quite nicely here.

Your force and velocity measurements have errors in them of various sorts. If your measurements were perfect you would only need one velocity measurement. The force (divided by mass) yields acceleration, and this integrates to velocity. Integrating a noisy measurement results in a random walk. A Kalman update with your velocity measurement provide a means to re-anchor that integrated velocity. Your velocity measurements are sporadic. That's not a problem; Kalman filters work quite nicely with sporadic updates.
 
  • Like
Likes berkeman
D H said:
An estimation technique such as a Kalman filter would work quite nicely here.

Your force and velocity measurements have errors in them of various sorts. If your measurements were perfect you would only need one velocity measurement. The force (divided by mass) yields acceleration, and this integrates to velocity. Integrating a noisy measurement results in a random walk. A Kalman update with your velocity measurement provide a means to re-anchor that integrated velocity. Your velocity measurements are sporadic. That's not a problem; Kalman filters work quite nicely with sporadic updates.
Unfortunately, the two variables do not relate to each other like that. In fact it is just their relationship that I am trying to find.
 
vlookup is a useful function, you can use the time column in continuous time series to look up matching values in the column with sporadic values. Make sure to set the range_lookup parameter to false

=VLOOKUP([cell in column of continuous data],[cell range for column of sporadic data and column of corresponding value],2 (gets the corresponding value),FALSE)

would need to be combined with the iferror() function if you are calculating something off the data
 
Learn If you want to write code for Python Machine learning, AI Statistics/data analysis Scientific research Web application servers Some microcontrollers JavaScript/Node JS/TypeScript Web sites Web application servers C# Games (Unity) Consumer applications (Windows) Business applications C++ Games (Unreal Engine) Operating systems, device drivers Microcontrollers/embedded systems Consumer applications (Linux) Some more tips: Do not learn C++ (or any other dialect of C) as a...

Similar threads

  • · Replies 15 ·
Replies
15
Views
4K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
17
Views
5K
  • · Replies 40 ·
2
Replies
40
Views
5K
Replies
10
Views
1K
  • · Replies 14 ·
Replies
14
Views
2K
Replies
2
Views
512
  • · Replies 1 ·
Replies
1
Views
4K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K