360 363 290 304 295 bytes
See at the bottom of the post for how to test the old code with Octave.
This code takes the name of the element (including Kalium, etc.) and displyas the output in ascii format now that the rules have changed.
f=input('');e=1;a=['CPACxxSAMSNxxxxxBLHxCKACSPSAMNNFONCBBLHH';'aorhxxilaoexxxxxeiexa rl ilgae eie '];for s=a;n=s(s~=32);if strncmpi(n,f,nnz(n));break;end;e=mod(e,20)+1;end;s=spiral(10);p=[8,18,33,28,23,39,60,53,46,95];p=[p;p+1];o=s*0;o(ismember(s,p(1:21-e)))='x';o(45:46)=a(:,e+20);char(o')
The rules changed since I wrote the code to require an ASCII output. I have updated my code to do this at the expense of 14 bytes. I have saved 9 bytes by getting rid of the reshape() and just making the
a matrix the right shape to begin with.
Here is an explanation of how it works:
%Get the name - actually we only need at most the first two characters, but the whole thing will do
%This bit makes a map which allows us to find the element (including with
%the names like Kalium. All of the elements appear twice, with the actual
%symbols being the second set. The first set gets all those whose names are
%either more than one character, or don't begin with the first two
%characters of the short for (e.g. Sodium). The string is reshaped into a
%2x40 array. 'Natrium' is a pain in the neck as it as it would get caught
%by 'N' for 'Nitrogen'. I have reversed the element order - so that all the
%ones beginning with N come before N. Some maths is done later on to
%correct for the number of electrons - basically 21-e so 1 becomes 20.
a=['CPACxxSAMSNxxxxxBLHxCKACSPSAMNNFONCBBLHH';'aorhxxilaoexxxxxeiexa rl ilgae eie '];
%For each group of 2 in the array of elements
%Remove any spaces from the name
%Do a comparison of the first one or two characters of the requested string
%break once the element is found
%If not this element add another electron. We wrap around after 20 as there are two copies of each
%e is now number of electrons
%Generate an array of points for each electron
%make an output array
%Plot all the points in is up to and including the number of electrons (see the notes above for why 21-e)
%And add the text in the centre - we extract the element name from the second group appearance in the 'a' array, hence adding 20.
%Display the result
This is the output for Hydrogen (ignore the dots, they are to avoid the lines being removed when showing here):
And here is the output for Calcium.
And the output for Natrium, which now works properly (before Natrium it would result in Nitrogen!).
xxNa x .
xx x .
The new version of the code doesn't work with Octave as it uses
spiral() which is only present in MATLAB.
You can however test the old code using the Octave online interpreter:
f=input('');e=1;a=['CPACxxSAMSNxxxxxBLHxCKACSPSAMNNFONCBBLHH';'aorhxxilaoexxxxxeiexa rl ilgae eie '];for s=a;n=s(s~=32);if strncmpi(n,f,nnz(n));break;end;e=mod(e,20)+1;end;u=14:(34-e);r=floor(u/8);t=u*pi/4;polar(t,r,'o');text(0,0,a(:,e+20)','horizontalalignment','c')
Run that, then enter a string like: 'Hydrogen' (including the quote marks). Once it is done, you will have to click the expand plot button (looks like a little graph symbol in the top right corner of the interpreter) to get it to show the full thing. In Octave it unfortunately add lines joining the points, this does not happen in MATLAB. But at least it allows you to test the logic behind it. As I say, this is still a graphical output, but you get the idea of how the elements are looked up.