The distribution pattern of raw data is uniform.
rawData =
RandomVariate[
TruncatedDistribution[{10, 50}, UniformDistribution[{28, 34}]],
100];
{FindDistribution@#, DistributionFitTest@#} &@rawData
How should it convert to the normal distribution?
I don’t think I’m really understanding what you want but here’s a possibility:
normalData = InverseCDF[NormalDistribution[0, 1], (rawData - 28)/(34 - 28)];
Or perhaps a PowerTransform?
Thanks~~ this is precisely the way I want!
I am not proficient in statistical methods~~
However, I know that keeping the data in the normal distribution is necessary for many statistical methods.
Manually adjusting the data distribution has been bothering me for a long time!
I was hoping to find a consistent way to put it into workflow!
The data is what is important. Forcing data to look normal to meet assumptions of a statistical procedure can be putting emphasis on the wrong thing. It’s maybe the statistical procedure that should be modified.