I am improving GauCh, a tool to evaluate the Gaussianity.

 What I did: 

I made four big changes.

1. Change the color bar of p-value from a continuous one to a discrete one. (Fig.1)
2. Make plots of "KS distance." (Fig.2)
3. Make plots with linear scale in frequencies. (Fig.3)
4. Remove known line noise (paint them white). (Fig.1~3)

 Detail explanation:

1. In terms of statistics, p-value is used only to determine whether rejects the null hypothesis or not.  Then, I set 0.05 as the significant value and change the color in the plot apparently by the threshold.
2. KS distance is a quantity used to calculate p-value. For the above reason, the size of p-value is meaningless. However, KS distance, which means the distance between two distributions (null hypothesis and data) can be considered as a value to evaluate the Gaussianity numerically.
3. So far, I made log scale ones only. Log scale ones should be used to see the behavior in lower frequencies, and linear ones should be in higher frequencies.
4. The line noise should be removed because it is already known to be non-Gaussian noise. I added an option to fill the area white when a list of center frequencies and widths of line noise is given. It is difficult to get the list, so now I used YamaT-san's result (Fig.4) for center frequencies and constant value, 0.3 Hz, for widths.