static void Main(string[] args)
{
var input1 = new Neural(new List <ISinapse <double> >(), (x, y) => 0);
var input2 = new Neural(new List <ISinapse <double> >(), (x, y) => 1);
var i1ToH1 = new Sinapse();
var i2ToH1 = new Sinapse();
Func <double, double, double> sum = (x, y) => x + y;
var h1 = new Neural(new List <ISinapse <double> >()
{
i1ToH1, i2ToH1
}, sum);
i1ToH1.Next = h1;
i2ToH1.Next = h1;
i1ToH1.Weight = 0.45;
i2ToH1.Weight = -0.12;
i1ToH1.Previous = input1;
i2ToH1.Previous = input2;
var i1ToH2 = new Sinapse();
var i2ToH2 = new Sinapse();
var h2 = new Neural(new List <ISinapse <double> >()
{
i1ToH2, i2ToH2
}, sum);
i1ToH2.Next = h2;
i2ToH2.Next = h2;
i1ToH2.Weight = 0.78;
i2ToH2.Weight = 0.13;
i1ToH2.Previous = input1;
i2ToH2.Previous = input2;
var h1ToO1 = new Sinapse();
var h2ToO1 = new Sinapse();
var o1 = new Neural(new List <ISinapse <double> > {
h1ToO1, h2ToO1
}, sum);
h1ToO1.Previous = h1;
h1ToO1.Weight = 1.5;
h2ToO1.Previous = h2;
h2ToO1.Weight = -2.3;
h1ToO1.Next = o1;
h2ToO1.Next = o1;
var res = o1.Calculate();
var error = GetError(res, 1);
Console.WriteLine($"result = {res} error = {error}");
}
private void getSamples()
{
try
{
Result = string.Empty;
List <Neural> firstMGroup = neurals.Where(ne => ne.W == T).ToList();
List <Neural> secondMGroup = neurals.Where(ne => ne.W > (double)T / 2 && ne.W < T).ToList();
List <Neural> thirdMGroup = neurals.Where(ne => ne.W == (double)T / 2).ToList();
List <Neural> fourthMGroup = neurals.Where(ne => ne.W < (double)T / 2).ToList();
List <Neural> firstMGroupStar = neurals.Where(ne => ne.W == T - 1).ToList();
List <Neural> secondMGroupStar = neurals.Where(ne => ne.W > (double)(T - 1) / 2 && ne.W < T - 1).ToList();
List <Neural> thirdMGroupStar = neurals.Where(ne => ne.W == (double)(T - 1) / 2).ToList();
List <Neural> fourthMGroupStar = neurals.Where(ne => ne.W < (double)(T - 1) / 2).ToList();
List <List <Neural> > result = new List <List <Neural> >();
//bool usedAllFourth = false;
//1 group
firstMGroup.ForEach(m => result.Add(new List <Neural> {
m
}));
//2 group
//int fourthCounter = 0;
if (fourthMGroup.Count == 0)
{
secondMGroup.ForEach(m => result.Add(new List <Neural> {
m
}));
}
else
{
foreach (var m in secondMGroup)
{
result.Add(new List <Neural> {
m, fourthMGroup.FirstOrDefault(f => f.W + m.W == T)
});
}
}
//3 group
if (thirdMGroup.Count % 2 == 0)
{
for (int i = 0; i < thirdMGroup.Count; i += 2)
{
result.Add(new List <Neural> {
thirdMGroup[i], thirdMGroup[i + 1]
});
}
}
else
{
for (int i = 0; i < thirdMGroup.Count - 1; i += 2)
{
result.Add(new List <Neural> {
thirdMGroup[i], thirdMGroup[i + 1]
});
}
var temp = new List <Neural> {
thirdMGroup[thirdMGroup.Count - 1]
};
if (fourthMGroup.Any())
{
Neural t = fourthMGroup.FirstOrDefault(f => f.W + thirdMGroup[thirdMGroup.Count - 1].W == T);
if (t != null)
{
temp.Add(t);
}
else
{
int i = 0;
while (temp.Sum(te => te.W) != T)
{
temp.Add(fourthMGroup[i++]);
}
}
}
result.Add(temp);
}
//4 group
List <Neural> neuralsForCheck = new List <Neural>();
foreach (var res in result)
{
neuralsForCheck.AddRange(res);
}
var fourthMGroupCopy = fourthMGroup.Where(g => !neuralsForCheck.Select(s => s.Number).Contains(g.Number)).ToList();
for (int i = 0; i < fourthMGroupCopy.Count; i++)
{
int res = fourthMGroupCopy[i].W;
List <Neural> temp = new List <Neural> {
fourthMGroupCopy[i]
};
if (res == T)
{
result.Add(temp);
break;
}
Neural t = fourthMGroup.FirstOrDefault(g => (g.W + res) == T && g.Number != fourthMGroupCopy[i].Number);
if (t != null)
{
temp.Add(t);
}
temp.Add(fourthMGroup.FirstOrDefault(g => g.Number != fourthMGroupCopy[i].Number));
res += fourthMGroup.FirstOrDefault(g => g.Number != fourthMGroupCopy[i].Number).W;
if (res == T)
{
result.Add(temp);
break;
}
List <Neural> tempNeurals = new List <Neural>();
tempNeurals.AddRange(firstMGroup);
tempNeurals.AddRange(secondMGroup);
tempNeurals.AddRange(thirdMGroup);
tempNeurals.AddRange(fourthMGroup);
foreach (var tempNeural in tempNeurals)
{
int a = res + tempNeural.W;
if (a == T)
{
temp.Add(tempNeural);
result.Add(temp);
break;
}
}
}
//STAR
List <List <Neural> > result2 = new List <List <Neural> >();
//1 group
firstMGroupStar.ForEach(m => result2.Add(new List <Neural> {
m
}));
//2 group
//int fourthCounter = 0;
if (fourthMGroupStar.Count == 0)
{
secondMGroupStar.ForEach(m => result2.Add(new List <Neural> {
m
}));
}
else
{
foreach (var m in secondMGroupStar)
{
result2.Add(new List <Neural> {
m, fourthMGroupStar.FirstOrDefault(f => f.W + m.W == T - 1)
});
}
}
//3 group
if (thirdMGroupStar.Count % 2 == 0)
{
for (int i = 0; i < thirdMGroupStar.Count; i += 2)
{
result2.Add(new List <Neural> {
thirdMGroupStar[i], thirdMGroupStar[i + 1]
});
}
}
else
{
for (int i = 0; i < thirdMGroupStar.Count - 1; i += 2)
{
result2.Add(new List <Neural> {
thirdMGroupStar[i], thirdMGroupStar[i + 1]
});
}
var temp = new List <Neural> {
thirdMGroupStar[thirdMGroupStar.Count - 1]
};
if (fourthMGroupStar.Any())
{
Neural t = fourthMGroupStar.FirstOrDefault(f => f.W + thirdMGroupStar[thirdMGroupStar.Count - 1].W == T - 1);
if (t != null)
{
temp.Add(t);
}
else
{
int i = 0;
while (temp.Sum(te => te.W) != T - 1)
{
temp.Add(fourthMGroup[i++]);
}
}
}
result2.Add(temp);
}
//4 group
neuralsForCheck = new List <Neural>();
foreach (var res in result2)
{
neuralsForCheck.AddRange(res);
}
fourthMGroupCopy = fourthMGroupStar.Where(g => !neuralsForCheck.Select(s => s.Number).Contains(g.Number)).ToList();
for (int i = 0; i < fourthMGroupCopy.Count; i++)
{
int res = fourthMGroupCopy[i].W;
List <Neural> temp = new List <Neural> {
fourthMGroupCopy[i]
};
if (res == T - 1)
{
result2.Add(temp);
break;
}
Neural t = fourthMGroupStar.FirstOrDefault(g => (g.W + res) == T - 1 && g.Number != fourthMGroupCopy[i].Number);
if (t != null)
{
temp.Add(t);
}
temp.Add(fourthMGroupStar.FirstOrDefault(g => g.Number != fourthMGroupCopy[i].Number));
res += fourthMGroupStar.FirstOrDefault(g => g.Number != fourthMGroupCopy[i].Number).W;
if (res == T - 1)
{
result2.Add(temp);
break;
}
List <Neural> tempNeurals = new List <Neural>();
tempNeurals.AddRange(firstMGroupStar);
tempNeurals.AddRange(secondMGroupStar);
tempNeurals.AddRange(thirdMGroupStar);
tempNeurals.AddRange(fourthMGroupStar);
foreach (var tempNeural in tempNeurals)
{
int a = res + tempNeural.W;
if (a == T - 1)
{
temp.Add(tempNeural);
result2.Add(temp);
break;
}
}
}
//checkForTable(result, result2);
//checkForCapacity(secondMGroup, fourthMGroup, firstMGroup.Count, thirdMGroup.Count, result);
//checkForCapacity2(secondMGroupStar, fourthMGroupStar, firstMGroupStar.Count, thirdMGroupStar.Count, result2);
//string s = string.Empty;
//foreach (var res in result)
//{
// int temp = res.Sum(r => r.W * Convert.ToInt32(r.X));
// s += (temp == T) + Environment.NewLine;
//}
////MessageBox.Show(s);
//Result += s + Environment.NewLine;
//s = string.Empty;
//foreach (var res in result2)
//{
// int temp = res.Sum(r => r.W * Convert.ToInt32(r.X));
// s += (temp == T - 1) + Environment.NewLine;
//}
////MessageBox.Show(s);
//Result += s + Environment.NewLine;
foreach (var res in result)
{
for (int i = 1; i <= N; i++)
{
if (!res.Select(r => r.Number).Contains(i))
{
Result += string.Format("¬X{0} ", i);
}
else
{
Result += string.Format("X{0} ", i);
}
}
Result += Environment.NewLine;
}
Result += Environment.NewLine;
foreach (var res in result2)
{
for (int i = 1; i <= N; i++)
{
if (!res.Select(r => r.Number).Contains(i))
{
Result += string.Format("¬X{0} ", i);
}
else
{
Result += string.Format("X{0} ", i);
}
}
Result += Environment.NewLine;
}
OnPropertyChanged("Result");
}
catch (Exception e)
{
MessageBox.Show(e.Message);
throw;
}
}