Standard Deviation calculattion

Share icon

Thu, 08/21/2014 - 03:41 am

Could someone tell me how Igor Pro can calculate the SD from values given in a wave? For example a certain quantity has been measured more than once and we have 2 values for it e.g. 3.15 and 3.28. How the SD can be calculated for these 2 values using Igor Pro (similarly to Excel).

Thank you in advance
Answer to how to calculate standard deviation in Igor Pro. If your values are in wave0 execute wavestats wave0. Standard deviation is stored in the variable V_sdev. In answer to whether you should do this for just two values... I cannot say.

August 21, 2014 at 04:35 am - Permalink

While a standard deviation from two values can be calculated (just like an "average" can be calculated for 1 value), it's fairly meaningless. Standard deviation, conceptually, is where ~68% of the data will lie if the data are Gaussian-distributed. So you need several data points to build up the statistics to understand where that ~68% envelope is. So you can do it, it just won't really mean much.

August 21, 2014 at 12:38 pm - Permalink

Igor also has a quick way to see the standard deviation of a wave in the data browser. On the top right of the data browser, there is a field to change some display options. Make sure the 'info' check-box is selected. There is a small 'i' box which, when clicked, can be switched to a sum symbol. In this state, the bottom text field shows various calculations (a bit like instant wavestats) for the currently selected wave, including the standard deviation.

August 21, 2014 at 07:13 pm - Permalink

Is there a simple command, such as
print mean(my wave)

to compute the standard deviation and for instance attribute this value to a variable su ch:
myvariable=mean(my wave)


thank you

April 10, 2017 at 02:23 am - Permalink

tijead wrote:
Is there a simple command, such as
print mean(my wave)

to compute the standard deviation and for instance attribute this value to a variable su ch:
myvariable=mean(my wave)


thank you


You can use Variance

For example, try:
make/O/D/N=1000 myWave=gnoise(2)
print sqrt(variance(myWave))


Hope this helps,
Kurt

April 10, 2017 at 03:39 am - Permalink