I'm doing some statistical programming in GAUSS 13, and I'm using it to, among other things, graph histogram plots of several large data series' distributions. I compute these distributions' "95% CI," and I want to draw vertical lines representing the 2.5 and 97.5 percentiles, i.e. the lower and upper bounds of these confidence intervals. I thought the natural way to do this would be to use _pline and just use the 2.5 and 97.5 percentile variable names as my x-coordinate starting and ending values; however, I discovered that when I use these variable names in _pline, _pline interprets them as denormal numbers (i.e. +DEN) rather than their actual values (which are all between -2 and 6--and are all significantly different from zero--which I checked many, many times in the command prompt). Why on Earth isn't this working?

## 1 Answer

0

I am not certain that I understand what you are doing. Please forgive me if my interpretation is in error. It sounds like you might be assigning a string value to the `_pline` global variable. However, the `_pline` global variable only takes numerical input.

Here is an example that draws a percentage histogram and adds a vertical line at -2:

library pgraph; _pline = { 1 6 -2 0 -2 15 0 3 10 }; x = rndn(1e4, 1); histp(x, 20);

But, since you have GAUSS 13 you could use the new graphics and just do this:

x = rndn(1e4, 1); line_x = { -2, -2 }; line_y = { 0, 0.18 }; //draw histogram plotHistP(x, 20); //add vertical line plotAddXY(line_x, line_y);

Then you could also use the interactive graphics editing to add annotations, arrows, change colors, etc.

## Your Answer

## 1 Answer

I am not certain that I understand what you are doing. Please forgive me if my interpretation is in error. It sounds like you might be assigning a string value to the `_pline` global variable. However, the `_pline` global variable only takes numerical input.

Here is an example that draws a percentage histogram and adds a vertical line at -2:

library pgraph; _pline = { 1 6 -2 0 -2 15 0 3 10 }; x = rndn(1e4, 1); histp(x, 20);

But, since you have GAUSS 13 you could use the new graphics and just do this:

x = rndn(1e4, 1); line_x = { -2, -2 }; line_y = { 0, 0.18 }; //draw histogram plotHistP(x, 20); //add vertical line plotAddXY(line_x, line_y);

Then you could also use the interactive graphics editing to add annotations, arrows, change colors, etc.