Showing numbers in decimal not scientific notation

tomy

@Chris-Kawa said in Showing numbers in decimal not scientific notation:

QString s;
QTextStream (&s) << QString::number(12.0, 'f');
lineEdit -> setText(s);

I used it. It's fine for doubles but when I calculate ints (e.g., 2+3) it shows 5.000000!!
We should make it show the precision only when the result is a double number not an int.

kshegunov

@tomy said in Showing numbers in decimal not scientific notation:

I used it. It's fine for doubles but when I calculate ints (e.g., 2+3) it shows 5.000000!!
We should make it show the precision only when the result is a double number not an int.

You either calculate a double or an int, can't be both. Anyway, what you probably want is this:

double calculationResult = 12.0;
QString displayValue = qFuzzyCompare(calculationResult, static_cast<qint64>(calculationResult)) ? QString::number(static_cast<qint64>(calculationResult)) : QString::number(calculationResult, 'f');

tomy

@kshegunov

You either calculate a double or an int, can't be both.
Anyway, what you probably want is this:

How can it be acceptable!? All calculator around the world do calculations on both types. Furthermore, when I write 2.3+4.6, it shows: 6.900000!
That is it works fine neither for ints nor for doubles.

double calculationResult = 12.0;
QString displayValue = qFuzzyCompare(calculationResult, static_cast<qint64>(calculationResult) ? QString::number(static_cast<qint64>(calculationResult)) : QString::number(calculationResult, 'f');

Ow my God! Isn't there any simpler way?
Please have a look at Windows built-in calculator. See this how simple and nifty shows results.
Do you say that behind that Windows' calculator there would be such a long statement just for showing numbers in decimal mode, if it were written by C++/Qt?

kshegunov

@tomy said in Showing numbers in decimal not scientific notation:

How can it be acceptable!? All calculator around the world do calculations on both types. Furthermore, when I write 2.3+4.6, it shows: 6.900000!

No they do calculation in the widest possible type they support (here it's double) and then display the result as appropriate. Further reading on implicit type promotions in c++ can be found here

Ow my God! Isn't there any simpler way?

This is a simple if-else statement with the notable exception that it compares floating point values as they should be compared.

Do you say that behind that Windows' calculator there would be such a long statement just for showing numbers in decimal mode, if it were written by C++/Qt?

Yes, I'm sure of it. It's probably even much longer as windows is actually written in C.

tomy

I used this:

ss = qFuzzyCompare(expression(),
                 static_cast<qint64>(expression()) ?
                 QString::number(static_cast<qint64>(expression())) :
                 QString::number(expression(), 'f'));
 result_box -> setText(ss);

ss is a QString.
expression() returns an int or double value.
result_box is a lineEdit which shows the result of the calculations.

I get this error:
C:\Users\ME\Documents\Qt\My_First_Calculator\my_first_calculator.cpp:81: error: no matching function for call to 'qFuzzyCompare(double, QString)'
QString::number(expression(), 'f'));
^

kshegunov

@tomy said in Showing numbers in decimal not scientific notation:

expression() returns an int or double value.

I'm pretty sure that functions returns a string, not an int, nor a double. You need to convert the string to an actual number if you want to use it as such, e.g. see here.

tomy

@kshegunov

expression() returns an int or double value.

I'm pretty sure that functions returns a string, not an int, nor a double.

Don't be that sure. :)
I have this method in my code:

double My_First_Calculator::expression()

But I think this error is of that ss is a QString.

VRonin

@tomy Did you even try my solution?

tomy

@VRonin

Should I use it this way:

QString ss;
ss = result_box -> locale().toString(1257.147862,'f');
QTextStream (&ss) << expression();
result_box -> setText(ss);

?

VRonin

@tomy No, don't use QTextStream you don't need something like stringstream to pass numbers to string.

QString ss;
ss = result_box -> locale().toString(expression(),'f');
result_box -> setText(ss);

or more concisely, result_box->setText(result_box->locale().toString(expression(),'f'));

tomy

@VRonin said in Showing numbers in decimal not scientific notation:

@tomy No, don't use QTextStream you don't need something like stringstream to pass numbers to string.
QString ss;
ss = result_box -> locale().toString(expression(),'f');
result_box -> setText(ss);

I used it. 2 + 3 = 5.000000 :( :(

VRonin

what is result_box and what is inside expression()?

tomy

@VRonin

what is result_box

It's a lineEdit.

and what is inside expression()?

It returns only a double value. Consider something simple like:

double My_First_Calculator::expression()
{
   double d1, d2;    // these d1, d2 are gotten from input E.g. d1 = 2, d2 = 3.5
  if(_ch == '+') return d1+d2;  // _ch is a previously defined varible 
 else if (_ch == '-') return d1-d2;  // and so on
}

tomy

I simplified the code as follows. This, too, has exactly that problem:

test.h

#ifndef TEST_H
#define TEST_H
#include <QDialog>

class QLineEdit;
class QPushButton;

class test : public QDialog
{
    Q_OBJECT
public:
    test(QWidget* parent = 0);

private slots:
    void expression();

private:
    QLineEdit* result_box;
    QPushButton* equal;
    QPushButton* quit;
};

#endif // TEST_H

test.cpp

#include <QtWidgets>
#include "test.h"

test::test(QWidget* parent) : QDialog(parent)
{
  result_box = new QLineEdit;
  equal = new QPushButton(tr("="));
  quit = new QPushButton(tr("Close"));


  connect(quit, SIGNAL(clicked(bool)), this, SLOT(close()));
  connect(equal,SIGNAL(clicked(bool)), this, SLOT(expression()));

  QHBoxLayout* layout = new QHBoxLayout;
  layout -> addWidget(result_box);
  layout -> addWidget(equal);
  layout -> addWidget(quit);

  setLayout(layout);
}

//******************

void test::expression()
{
    QString ss;
    double d = 1000000;
    QTextStream (&ss) << d;
    result_box -> setText(ss);
}

And main.cpp

#include <QApplication>
#include "test.h"

int main(int argc, char* argv[])
{
    QApplication app(argc, argv);
    test t;
    t.show();

   return app.exec();
}

Just run it and click on the = button.

tomy

I think I should write a code for it like this:

if( d is like an int number)
    result_box -> setText(QString::number(d , 'f', 0));
else if ( d is a double number with n numbers after point)
   result_box -> setText(QString::number(d , 'f', n));

koahnig

@tomy
Maybe you should have a look to the docs as well. E.g. here

In case of integer assignment that would be:

int i = 10;
result_box -> setText ( QString::number( i ) );

kshegunov

@tomy said in Showing numbers in decimal not scientific notation:

Don't be that sure. :)

Yes! There's a typo in the code ... :)

ss = qFuzzyCompare(expression(),
                 static_cast<qint64>(expression()) ? //< Missing a )
                 QString::number(static_cast<qint64>(expression())) :
                 QString::number(expression(), 'f')); //< Extra )

What you want is to have the if with qFuzzyCompare, not with the static cast. It should rather read like this:

ss = qFuzzyCompare(expression(), static_cast<qint64>(expression())) ?
                 QString::number(static_cast<qint64>(expression())) :
                 QString::number(expression(), 'f');

If you wish you can of course use the usual if-else construct, not the ternary operator, so the last snippet'd be equivalent to:

if (qFuzzyCompare(expression(), static_cast<qint64>(expression()))
    ss = QString::number(static_cast<qint64>(expression()));
else
    ss = QString::number(expression(), 'f');

tomy

This post is deleted!

tomy

@kshegunov

What you want is to have the if with qFuzzyCompare, not with the static cast. It should rather read like this:
ss = qFuzzyCompare(expression(), static_cast<qint64>(expression())) ?
                 QString::number(static_cast<qint64>(expression())) :
                 QString::number(expression(), 'f');
If you wish you can of course use the usual if-else construct, not the ternary operator, so the last snippet'd be equivalent to:
if (qFuzzyCompare(expression(), static_cast<qint64>(expression()))
    ss = QString::number(static_cast<qint64>(expression()));
else
    ss = QString::number(expression(), 'f');

Sorry, I don't know how I tested it but it doesn't work as it's expected! :(
for example:
2 + 3 = 5 OK
10 ^ 6 = 1000000 OK
2 + 1.5 = 3.500000 !!!

That is, the function QString::number(expression(), 'f'); (with the default precision 6) shows all range of precision whether it's need or not!

Maybe it's what we need:
http://doc.qt.io/qt-5/qtextstream.html#setRealNumberNotation
But how to use it? Docs doesn't offer a little example of it to show how beginners should use it. !!!!!!
(Docs are not for learners/beginners, they are for professionals — if they are professional, so they don't need Docs much => Docs are not useful)

kshegunov

What you want to do in the floating point case is not at all trivial. Here you can read on that topic if you are willing to digest the math.

The simpler but inefficient approach is to save the floating point text representation and just discard all the trailing 0 characters manually.

EDIT:
Here's (probably) the most concise way, but requires knowledge of regular expressions, which is yet again for you to read on to understand how or why it works:

if (qFuzzyCompare(expression(), static_cast<qint64>(expression()))
    ss = QString::number(static_cast<qint64>(expression()));
else  {
    ss = QString::number(expression(), 'f', 17);
    ss.replace(QRegularExpression("\\.?0+$"), "");
}

(Docs are not for learners/beginners, they are for professionals — if they are professional, so they don't need Docs much => Docs are not useful)

You got that backwards. The point of the Qt docs is they are a documentation for Qt itself, they will give no instruction into C++, its types, the types' memory layout and other such technical topics. If you are in need of that, then you need to look elsewhere, for example http://en.cppreference.com/w/