Hi.
what does gaussian distribution histograms measure? like what is the distance that gaussian measure? and what do the figures in red tells me when you keep moving them different percentages appear
thanks
Gaussian distribution
Gaussian distribution
- Attachments
-
- Capture.PNG (27.05 KiB) Viewed 4846 times
Re: Gaussian distribution
Another question, how to plot standard deviation data in Excel? I have exported histogram to Exel via the options top right corner on the histogram, but I'm not sure of the data. There are class, value, class start and class end. Which of these data to plot? thanks
- Attachments
-
- Capture std.PNG (180.49 KiB) Viewed 4838 times
Re: Gaussian distribution
1) Well, the 'Compute Statistical Parameters' tool simply fit a statistical distribution (here a Normal/Gaussian distribution) to your scalar field:
http://www.cloudcompare.org/doc/wiki/in ... parameters
To be meaningful, it's better if your data follows such a distribution (which is not the case here, but that's not too bad). At least you can get the average and standard deviation of your scalar field values.
The vertical line is not related to the distribution (it's actually displayed with any histogram view in CC). It simply displays the value and the number of samples below the line position, as well as the percentage of values on the left of the line.
2) From the histogram values, you can't compute the standard deviation (you need all the values to do it accurately). Exporting the histogram values is only useful to draw the same histogram but with Excel ;)
http://www.cloudcompare.org/doc/wiki/in ... parameters
To be meaningful, it's better if your data follows such a distribution (which is not the case here, but that's not too bad). At least you can get the average and standard deviation of your scalar field values.
The vertical line is not related to the distribution (it's actually displayed with any histogram view in CC). It simply displays the value and the number of samples below the line position, as well as the percentage of values on the left of the line.
2) From the histogram values, you can't compute the standard deviation (you need all the values to do it accurately). Exporting the histogram values is only useful to draw the same histogram but with Excel ;)
Daniel, CloudCompare admin
Re: Gaussian distribution
I have plotted the standard deviation data (std cloud 1 and std cloud 2) generated from M3C2 in excel in form of bar chart. I have taken the averages of both, but I got different values than the histogram. For example, the average of std cloud 2 is 0.052 but the histogram value is 0.032. which value is more accurate and what is std value in the histogram? thanks
- Attachments
-
- std excel.PNG (111.35 KiB) Viewed 4789 times
-
- std averages.PNG (11.2 KiB) Viewed 4789 times
-
- std histogram.PNG (24.23 KiB) Viewed 4789 times
Re: Gaussian distribution
The standard deviation of a set of values is not its average (you have to call 'STDEV' on Excel to compute it). But I guess here you compare the Gauss 'mean" value with the average of the values.
Pay attention that this is the 'best fit' Gauss/Normal distribution here. It's not necessarily the same as the average and standard deviation of the set of values. If your data doesn't follow a Normal distribution (i.e. the Chi2 distance is too high), then you'd better keep the average of the values (depending on what you are looking for of course ;)
Pay attention that this is the 'best fit' Gauss/Normal distribution here. It's not necessarily the same as the average and standard deviation of the set of values. If your data doesn't follow a Normal distribution (i.e. the Chi2 distance is too high), then you'd better keep the average of the values (depending on what you are looking for of course ;)
Daniel, CloudCompare admin
Re: Gaussian distribution
ah alright. So the mean should be the same as the average of the value. I have just used STEV of the values and I have the same results as the one on the histogram.
You mean if my data doesn't follow a normal distribution, the mean could be different than the average of std cloud 1 (for example). if the mean from the histogram is 0.037968 and the average of std cloud 1 (from excel) is 0.05175067. What does this tell me?
You mean if my data doesn't follow a normal distribution, the mean could be different than the average of std cloud 1 (for example). if the mean from the histogram is 0.037968 and the average of std cloud 1 (from excel) is 0.05175067. What does this tell me?