Reputation: 157
I have this source code in Delphi, why I get this error "Floating point overflow." when I run the code? and how to correct it?
The error message:
The code:
procedure TForm1.Button1Click(Sender: TObject);
var n, d, i, j, maxiter , iter: Integer;
Lower,Upper : Double;
X, V : TArray<TArray<Double>>;
begin
Lower := 0;
Upper := 0.2;
n := 100;
d := 55;
SetLength(V, n, d);
SetLength(X, n, d);
maxiter := 2000;
iter := 1;
for i:= 0 n-1 do
for j:=0 to d-1 do
begin
X[i][j]:= Lower + (Upper - Lower) * Random;
V[i][j] := 0.1 * X[i][j];
end;
while (iter <= maxiter) do
begin
for i:= 0 to n-1 do
for j:= 0 to D-1 do
V[i][j]:= 5 * V[i][j] + 2.0 * Random;
iter := iter +1;
end;
end;
Upvotes: 3
Views: 3262
Reputation: 80177
Look here: V[i][j]:= 5 * V[i][j] + 2.0 * Random;
You make 2000 iterations, so your results might be as large as 7^2000 ~ 10^1690, but max value for Double type is about 10^308. So “Floating point overflow” error is exact diagnosis.
You could see V[] values about 10^307 in debug watch or immediate watch (mouse over V[]) when error occurred.
You can use 10-byte Extended type(probably not available for 64-bit compilers) to avoid overflow for these given variable values, but this is not good solution in general case.
Aside note: You did not set i index value for this code piece:
for j:=0 to d-1 do
begin
X[i][j]:= Lower + (Upper - Lower) * Random;
V[i][j] := 0.1 * X[i][j];
end;
Upvotes: 10