Custom table model save only six characters on float

zloi_templar

Hello, i have my own class CustomTableModel based on QAbstractTableModel

When i try to save some numbers, for exameple 111111.11 he save only 111111

bool CustomTableModel::setData(const QModelIndex &index, const QVariant &value, int role)
{
    if (index.isValid() && role == Qt::EditRole)
    {
        Columns[index.row()][index.column()] = value.toFloat();
        qDebug()<<value.toString()<<value.toFloat();
        emit(dataChanged(index, index));
        return true;
    }
    return false;
}

If i made Columns type QVector<QVector<QString>> Columns, all works correct, but i don't wanna work with string and convertations. Can i save correct float values without losing decimal places?
Any suggestions?

Christian Ehrlicher

@zloi_templar said in Custom table model save only six characters on float:

When i try to save some numbers, for exameple 111111.11 he save only 111111

This is wrong - it only displays 6 digits I would guess. It's just a problem on how it is displayed - how did you check?

zloi_templar

@Christian-Ehrlicher
i save values on db, it's save only 6 digits

query.addBindValue(view->model()->data(view->model()->index(i, 0 , QModelIndex()), Qt::EditRole).toDouble());

if i tried

QVariant CustomTableModel::data(const QModelIndex &index, int role) const
{
if ((role == Qt::EditRole) && !std::isnan(Columns[index.row()][index.column()]))
        return QString::number(Columns[index.row()][index.column()], 'f', 6);
}

he print something like this:
qDebug()<<value.toString()<<QString::number(value.toFloat(), 'f', 6);
"123456.123456" "123456.125000"

but numbers with 7 and more letters before dot it save like:
"12345678.12345" "12345678.000000"

JonB

@zloi_templar
I don't understand:

Why are you "saving" (whatever you mean by that) a number as a string? You're asking for trouble, save it as a number.
Is using toFloat() a good idea? floats have limited precision, at least use toDouble()/doubles?
Even then you will lose accuracy. The only "safe" way to store accurate floating point numbers in a database is via its decimal type support.