Feature/so mouse movement + computer score calculation

[TimonKapischke]

No description provided.

[TimonKapischke]

I fixed all bugs.

[TimonKapischke]

I haven't tested the program yet. (I have problems with pytorch. The computer cant find a package.)

[felbecker]

Thanks.
Since scrollbars do not appear currently, can you change the default value of the flag to false such that your mouse movement is the default behavior?

[felbecker]

The view is incorrect for me after selecting the first edge. Enemy edge selection seems to work, however, there must be still a bug in rendering. Maybe something with the new enemy edges? @Luc4sQ

[TimonKapischke]

I changed the default value.

[MarioStanke]

Das Programm crasht beim anfänglichen Lernen

Epoch: 29 | Loss: 7683.76
Epoch: 30 | Loss: 7436.13
terminate called after throwing an instance of 'std::domain_error'
  what():  value cannot be converted to type float without overflow: 2.65835e+78
Aborted (core dumped)

Einstellungen: Standard, außer k=2

[MarioStanke]

Noch ein Bug im Training:

Epoch: 30 | Loss: 4.31349e+18
terminate called after throwing an instance of 'std::out_of_range'
  what():  vector<bool>::_M_range_check: __n (which is 4294967295) >= this->size() (which is 106)
Aborted (core dumped)

[felbecker]

Stimmt. Passiert aber nicht bei allen Instanzen. Der out_of_range Fehler tritt scheinbar auf, wenn das Training divergiert. Kleinere Lernrate, sodass es sicher für alle zufällig generierten Instanzen klappt? Ansonsten prüfen, ob loss ansteigt und Training notfalls abbrechen, um den Fehler zu vermeiden.