Least Squares Monte Carlo for European Options

Hi, I am trying to compare Spectral Methods, Black Scholes analytical and Least Squares Monte Carlo for European Put options.While Spectral method gives me a range of option values with different spot prices and its the same for Black Scholes. In case of LSM, I need to define a starting spot price and thus I am not getting the whole range of values like the other two methods. Can anyone help me out with a way to do least squares monte carlo and get a similar range like the spectral methods. Below is my LSM code.

T = 1.0;            % Time to maturity
r = 0.03;           % Risk free interest rate
sigma = 0.2;        % Volatility
K = 40;             % Strike price
S0 = 36;            % Underlying asset price in-the-money
%S0 = 44;           % Underlying asset price out-of-money
N = 50;             % number of time steps
M = 100000;         % 100k price paths
k = 3;              % Number of Basis Functions used for Laguerre Polyomials

dt = T/N;           % value of time step
t = 0:dt:T;         % time vector

% Generate stock price matrix (100k price paths) 

S = ones(M, N+1);    
R = exp((r - sigma^2/2)*dt+sigma*sqrt(dt)*randn(N,M));
SS = cumprod([S0*ones(1,M); R]);
S = SS';
P = max((K - S(:,N+1)),0);      %payoff at time T
for i = N:-1:2
  itmP = find(S(:,i));        % All paths
  X =  S(itmP,i);             % All price paths
  Y = P(itmP)*exp(-r*dt);     % discounted payoffs from continuation
  A = BasisFunctions(X,k);    % calculates basis functions
  beta = A\Y;                 % regression coeff
  C = A * beta;               % Estimated continuation value
  exP = itmP(K - X>0);        % Paths where strike greater than spot
  P(exP) = P(exP)*exp(-r*dt); %Insert payoffs and discount back one step

    u = mean(P * exp(-r*dt));   % Value of the option
    opt_val(i) = u;
    plot(X,C,'-m')