do the following matlab code using QPSK instead of BPSK:

clc; clear all; close all;

training_lenght=10^4;%the length of training data

training_data=randi([0,1],1,training_lenght);%generating random training data

qpsk=comm.BPSKModulator;%we used bpsk modulation

training_data=transpose(training_data);%transposed the data because qpsk takes a vertical matrix

training_data=qpsk(training_data);

training_data=transpose(training_data);%transposed it back to its original shape

h=[0.74 -0.514 0.37];%we chose a three tap channel

chan_len=length(h);

P=sum(abs(training_data).^2)/training_lenght;

snr=db2pow(10);

var=P/snr;

n=sqrt(var).*randn(1,training_lenght+2);%generated a random noise

y=conv(h,training_data)+n;%going through the channel

equalizer_length=30;

equalizer=zeros(1,equalizer_length);

Rvv0=(training_data*training_data')/(training_lenght+chan_len-1);%the power of the equalizer input, the output of the channel

max_step=2/(equalizer_length*Rvv0);%step size must be between this value and 0

step_size=max_step*0.0125;%I choose step size arbitrarily

for i1=1:training_lenght-equalizer_length+1

equalizer_ip = fliplr(y(i1:i1+equalizer_length-1));%equalizer input

error = training_data(i1+equalizer_length-1)- equalizer*equalizer_ip.';% instantaneous error

equalizer = equalizer + step_size*error*equalizer_ip;

end

message_length=10^5;

message=randi([0,1],1,message_length);%generating random data to transmit

message2=transpose(message);

mod_message=qpsk(message2);

mod_message=transpose(mod_message);

P=sum(abs(mod_message).^2)/length(mod_message);

var=P/snr;

n=sqrt(var).*randn(1,message_length+2);%generated a random noise

channel_output=conv(h,mod_message)+n;

equalizer_output=conv(equalizer,channel_output);

equalizer_output=equalizer_output(1:message_length);

qpskdemod = comm.BPSKDemodulator;

equalizer_output=transpose(equalizer_output);

y= qpskdemod(equalizer_output);

y=transpose(y);

BER=0;

for i=1:message_length

if y(i) ~= message(i)

BER=BER+1;

end

end

BER=BER/message_length;