59943_en-US

liliconcom

How to fit a normal distribution of specified mean and standard deviation to a data set?

Suppose I create 100 values from a standard normal.

While my parameters are mean=0 and SD=1, the calculated statistics from the 100 data will not be fully equal to the parameter values.

If I use "Analyze" and "Distribution" I get a histogram of the 100 random data values and I can fit a normal distribution to these values ("Continuous fit" / "Normal"). The fit is based on the statistics of the data.

Rather than fitting a normal using the statistics as estimates of the parameters, can I fit a normal distribution using the mean and SD parameters I specify? (In this case zero and one.)

Would be great to have some advice on this.

Thanks!

 

 

 

Re: How to fit a normal distribution of specified mean and standard deviation to a data set?Thank you, Mark.Re: How to fit a normal distribution of specified mean and standard deviation to a data set?

You are not estimating the parameters, you are specifying them. I think you want to plot the PDF for the normal distribution over the histogram. A simple customization is sufficient. If you view the histogram in the default vertical orientation, then use the X Function( Normal Density( x, 0, 1 ), x ); If you rotated the orientation then use the Y Function( Normal Density( x, 0, 1 ), x ); Right-click in the histogram frame box and select Customize. Click the + button and add the JSL.

You get extra points if you figure out how to determine the curvecolor, style, and width...