Problem with wrf.interplevel desiredlev input

Hello, I encountered a problem when trying to use a "for loop" to get a sequence of different interpolation planes.

So I have tried this line, which works fine:
p_certain_level = wrf.interplevel(p,z,2893.9905)

I can plot this to get:

I also have other different height values that I want to plot. So I write a list:
height = [588.0741, 1751.7858, 2893.9905, 4019.048, 5129.8022, 6228.1494, ... ]

Then I write the for loop:
for a in range(0,3): ## here i'm just trying the first three heights, not the whole list
ss = wrf.interplevel(p, z, float(height[a]))
print(ss)

This returns something like:

As you can see, there are a lot of nan in the printed array. I also tried to plot the result:

As you can see, the middle values are kept unchanged. However, the other values are all missing. I'm not sure what's happening that leads to interplevel method not working. I've tried a few times and believe this is related to list and numpy.array (both list and np.array are producing the same problem).

I've tried wrf.getvar, in which I use a "timeidx = blabla_list [3]", which turns out to work properly.

Start by reproducing the scalar and looped wrf.interplevel calls with the same p, z, and height values shown in the issue, then inspect how NaN values are produced for each requested level. No repository file or test is named, so compare the interpolation behavior for scalar, list, and numpy-array inputs. Done means determining whether the missing values are caused by the input values or by wrf.interplevel, with a documented reproduction or fix.