/// <summary>
/// Generates all valid binary combinations of all binary configurations options in the given model
/// </summary>
/// <param name="vm">The variability model containing the binary options and their constraints.</param>
/// <returns>Returns a list of configurations, in which a configuration is a list of SELECTED binary options (deselected options are not present)</returns>
public List <List <BinaryOption> > generateAllVariantsFast(VariabilityModel vm)
{
List <List <BinaryOption> > configurations = new List <List <BinaryOption> >();
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
ConstraintSolverSolution soln = S.Solve();
while (soln.HasFoundSolution)
{
List <BinaryOption> config = new List <BinaryOption>();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
config.Add(termToElem[cT]);
}
}
//THese should always be new configurations
// if(!Configuration.containsBinaryConfiguration(configurations, config))
configurations.Add(config);
soln.GetNext();
}
return(configurations);
}
/// <summary>
/// The method aims at finding a configuration which is similar to the given configuration, but does not contain the optionToBeRemoved. If further options need to be removed from the given configuration, they are outputed in removedElements.
/// Idea: Encode this as a CSP problem. We aim at finding a configuration that maximizes a goal. Each option of the given configuration gets a large value assigned. All other options of the variability model gets a negative value assigned.
/// We will further create a boolean constraint that forbids selecting the optionToBeRemoved. Now, we find an optimal valid configuration.
/// </summary>
/// <param name="optionToBeRemoved">The binary configuration option that must not be part of the new configuration.</param>
/// <param name="originalConfig">The configuration for which we want to find a similar one.</param>
/// <param name="removedElements">If further options need to be removed from the given configuration to build a valid configuration, they are outputed in this list.</param>
/// <param name="vm">The variability model containing all options and their constraints.</param>
/// <returns>A configuration that is valid, similar to the original configuration and does not contain the optionToBeRemoved. Null if not configuration could be found.</returns>
public List <BinaryOption> generateConfigWithoutOption(BinaryOption optionToBeRemoved, List <BinaryOption> originalConfig, out List <BinaryOption> removedElements, VariabilityModel vm)
{
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
removedElements = new List <BinaryOption>();
//Forbid the selection of this configuration option
CspTerm optionToRemove = elemToTerm[optionToBeRemoved];
S.AddConstraints(S.Implies(S.True, S.Not(optionToRemove)));
//Defining Goals
CspTerm[] finalGoals = new CspTerm[variables.Count];
int r = 0;
foreach (var term in variables)
{
if (originalConfig.Contains(termToElem[term]))
{
finalGoals[r] = term * -1000; //Since we minimize, we put a large negative value of an option that is within the original configuration to increase chances that the option gets selected again
}
else
{
finalGoals[r] = variables[r] * 10000;//Positive number will lead to a small chance that an option gets selected when it is not in the original configuration
}
r++;
}
S.TryAddMinimizationGoals(S.Sum(finalGoals));
ConstraintSolverSolution soln = S.Solve();
List <BinaryOption> tempConfig = new List <BinaryOption>();
if (soln.HasFoundSolution && soln.Quality == ConstraintSolverSolution.SolutionQuality.Optimal)
{
tempConfig.Clear();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
tempConfig.Add(termToElem[cT]);
}
}
//Adding the options that have been removed from the original configuration
foreach (var opt in originalConfig)
{
if (!tempConfig.Contains(opt))
{
removedElements.Add(opt);
}
}
return(tempConfig);
}
return(null);
}
/// <summary>
/// Checks whether the boolean selection is valid w.r.t. the variability model. Does not check for numeric options' correctness.
/// </summary>
/// <param name="config">The list of binary options that are SELECTED (only selected options must occur in the list).</param>
/// <param name="vm">The variability model that represents the context of the configuration.</param>
/// <param name="exact">Checks also the number of selected options such that it returns only true if exactly the given configuration is valid
/// (e.g., if we need to select more features to get a valid config, it returns false if exact is set to true).
/// <returns>True if it is a valid selection w.r.t. the VM, false otherwise</returns>
public bool checkConfigurationSAT(List <BinaryOption> config, VariabilityModel vm, bool exact)
{
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
//Feature Selection
foreach (BinaryOption binayOpt in elemToTerm.Keys)
{
CspTerm term = elemToTerm[binayOpt];
if (config.Contains(binayOpt))
{
S.AddConstraints(S.Implies(S.True, term));
}
else
{
if (exact)
{
S.AddConstraints(S.Implies(S.True, S.Not(term)));
}
}
}
ConstraintSolverSolution sol = S.Solve();
if (sol.HasFoundSolution)
{
int count = 0;
foreach (CspTerm cT in variables)
{
if (sol.GetIntegerValue(cT) == 1)
{
count++;
}
}
//Needs testing TODO
if (count != config.Count && exact == true)
{
return(false);
}
return(true);
}
else
{
return(false);
}
}
private List <List <BinaryOption> > generateTilSize(int i1, int size, int timeout, VariabilityModel vm)
{
var foundSolutions = new List <List <BinaryOption> >();
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
CspTerm t = S.ExactlyMofN(i1, variables.ToArray());
S.AddConstraints(new CspTerm[] { t });
var csp = new ConstraintSolverParams
{
TimeLimitMilliSec = timeout * 1000,
};
ConstraintSolverSolution soln = S.Solve(csp);
int counter = 0;
while (soln.HasFoundSolution)
{
List <BinaryOption> tempConfig = (
from cT
in variables
where soln.GetIntegerValue(cT) == 1
select termToElem[cT]).ToList();
if (tempConfig.Contains(null))
{
tempConfig.Remove(null);
}
foundSolutions.Add(tempConfig);
counter++;
if (counter == size)
{
break;
}
soln.GetNext();
}
//Console.WriteLine(i1 + "\t" + foundSolutions.Count);
return(foundSolutions);
}
/* public List<List<string>> generateDimacsPseudoRandom(string[] lines,int features, int randomsize, BackgroundWorker worker)
* {
* var cts = new CancellationTokenSource();
* var erglist = new List<List<string>>();
*
* var tasks = new Task[features];
* var mylock = new object();
*
* for (var i = 0; i < features; i++)
* {
* var i1 = i;
* tasks[i] = Task.Factory.StartNew(() =>
* {
* Console.WriteLine("Starting: " + i1);
* var sw = new Stopwatch();
* sw.Start();
* var result = generateDimacsSize(lines, i1, randomsize);
* sw.Stop();
* Console.WriteLine("Done: " + i1 + "\tDuration: " + sw.ElapsedMilliseconds + "\tResults: " + result.Count);
* return result;
*
* }, cts.Token).ContinueWith(task =>
* {
* lock (mylock)
* {
*
* erglist.AddRange(task.Result);
*
* Console.WriteLine("Added results: " + i1 + "\tResults now: " + erglist.Count);
* counter++;
* //worker.ReportProgress((int)(counter * 100.0f / (double)features), erglist.Count);
* }
* });
*
* if (Task.WaitAny(new[] { tasks[i] }, TimeSpan.FromMilliseconds(60 * 1000)) < 0)
* {
* cts.Cancel();
* }
* }
*
* Task.WaitAll(tasks);
*
* return erglist;
* } */
/// <summary>
/// Simulates a simple method to get valid configurations of binary options of a variability model. The randomness is simulated by the modulu value.
/// We take only the modulu'th configuration into the result set based on the CSP solvers output. If modulu is larger than the number of valid variants, the result set is empty.
/// </summary>
/// <param name="vm">The variability model containing the binary options and their constraints.</param>
/// <param name="treshold">Maximum number of configurations</param>
/// <param name="modulu">Each configuration that is % modulu == 0 is taken to the result set. Can be less than the maximal (i.e. threshold) specified number of configurations.</param>
/// <returns>Returns a list of configurations, in which a configuration is a list of SELECTED binary options (deselected options are not present</returns>
public List <List <BinaryOption> > generateRandomVariants(VariabilityModel vm, int treshold, int modulu)
{
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
List <List <BinaryOption> > erglist = new List <List <BinaryOption> >();
ConstraintSolverSolution soln = S.Solve();
int mod = 0;
while (soln.HasFoundSolution)
{
mod++;
if (mod % modulu != 0)
{
soln.GetNext();
continue;
}
List <BinaryOption> tempConfig = new List <BinaryOption>();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
tempConfig.Add(termToElem[cT]);
}
}
if (tempConfig.Contains(null))
{
tempConfig.Remove(null);
}
erglist.Add(tempConfig);
if (erglist.Count == treshold)
{
break;
}
soln.GetNext();
}
return(erglist);
}
public bool checkDimacsSAT(List <string> tempConfig, string[] lines)
{
var exact = false;
List <CspTerm> variables = new List <CspTerm>();
Dictionary <string, CspTerm> elemToTerm = new Dictionary <string, CspTerm>();
Dictionary <CspTerm, string> termToElem = new Dictionary <CspTerm, string>();
var sys = CSPsolver.getDimacsCSystem(out variables, out elemToTerm, out termToElem, lines);
foreach (var k in elemToTerm.Keys)
{
if (tempConfig.Contains(k))
{
sys.Implies(sys.True, elemToTerm[k]);
}
}
var sol = sys.Solve();
return(sol.HasFoundSolution);
}
/// <summary>
/// Based on a given (partial) configuration and a variability, we aim at finding all optimally maximal or minimal (in terms of selected binary options) configurations.
/// </summary>
/// <param name="config">The (partial) configuration which needs to be expaned to be valid.</param>
/// <param name="vm">Variability model containing all options and their constraints.</param>
/// <param name="minimize">If true, we search for the smallest (in terms of selected options) valid configuration. If false, we search for the largest one.</param>
/// <param name="unwantedOptions">Binary options that we do not want to become part of the configuration. Might be part if there is no other valid configuration without them</param>
/// <returns>A list of configurations that satisfies the VM and the goal (or null if there is none).</returns>
public List <List <BinaryOption> > maximizeConfig(List <BinaryOption> config, VariabilityModel vm, bool minimize, List <BinaryOption> unwantedOptions)
{
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
//Feature Selection
if (config != null)
{
foreach (BinaryOption binOpt in config)
{
CspTerm term = elemToTerm[binOpt];
S.AddConstraints(S.Implies(S.True, term));
}
}
//Defining Goals
CspTerm[] finalGoals = new CspTerm[variables.Count];
for (int r = 0; r < variables.Count; r++)
{
if (minimize == true)
{
BinaryOption binOpt = termToElem[variables[r]];
if (unwantedOptions != null && (unwantedOptions.Contains(binOpt) && !config.Contains(binOpt)))
{
finalGoals[r] = variables[r] * 10000;
}
else
{
// Element is part of an altnerative Group ... we want to select always the same option of the group, so we give different weights to the member of the group
//Functionality deactivated... todo needs further handling
/*if (binOpt.getAlternatives().Count != 0)
* {
* finalGoals[r] = variables[r] * (binOpt.getID() * 10);
* }
* else
* {*/
finalGoals[r] = variables[r] * 1;
//}
// wenn in einer alternative, dann bekommt es einen wert nach seiner reihenfolge
// id mal 10
}
}
else
{
finalGoals[r] = variables[r] * -1; // dynamic cost map
}
}
S.TryAddMinimizationGoals(S.Sum(finalGoals));
ConstraintSolverSolution soln = S.Solve();
List <string> erg2 = new List <string>();
List <BinaryOption> tempConfig = new List <BinaryOption>();
List <List <BinaryOption> > resultConfigs = new List <List <BinaryOption> >();
while (soln.HasFoundSolution && soln.Quality == ConstraintSolverSolution.SolutionQuality.Optimal)
{
tempConfig.Clear();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
tempConfig.Add(termToElem[cT]);
}
}
if (minimize && tempConfig != null)
{
resultConfigs.Add(tempConfig);
break;
}
if (!Configuration.containsBinaryConfiguration(resultConfigs, tempConfig))
{
resultConfigs.Add(tempConfig);
}
soln.GetNext();
}
return(resultConfigs);
}
/// <summary>
/// Based on a given (partial) configuration and a variability, we aim at finding the smallest (or largest if minimize == false) valid configuration that has all options.
/// </summary>
/// <param name="config">The (partial) configuration which needs to be expaned to be valid.</param>
/// <param name="vm">Variability model containing all options and their constraints</param>
/// <param name="minimize">If true, we search for the smallest (in terms of selected options) valid configuration. If false, we search for the largest one.</param>
/// <param name="unWantedOptions">Binary options that we do not want to become part of the configuration. Might be part if there is no other valid configuration without them.</param>
/// <returns>The valid configuration (or null if there is none) that satisfies the VM and the goal.</returns>
public List <BinaryOption> minimizeConfig(List <BinaryOption> config, VariabilityModel vm, bool minimize, List <BinaryOption> unWantedFeatures)
{
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
//Feature Selection
foreach (BinaryOption binOpt in config)
{
CspTerm term = elemToTerm[binOpt];
S.AddConstraints(S.Implies(S.True, term));
}
//Defining Goals
CspTerm[] finalGoals = new CspTerm[variables.Count];
for (int r = 0; r < variables.Count; r++)
{
if (minimize == true)
{
if (unWantedFeatures != null && (unWantedFeatures.Contains(termToElem[variables[r]]) && !config.Contains(termToElem[variables[r]])))
{
finalGoals[r] = variables[r] * 100;
}
else
{
finalGoals[r] = variables[r] * 1;
}
}
else
{
finalGoals[r] = variables[r] * -1; // dynamic cost map
}
}
S.TryAddMinimizationGoals(S.Sum(finalGoals));
ConstraintSolverSolution soln = S.Solve();
List <string> erg2 = new List <string>();
List <BinaryOption> tempConfig = new List <BinaryOption>();
while (soln.HasFoundSolution)
{
tempConfig.Clear();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
tempConfig.Add(termToElem[cT]);
}
}
if (minimize && tempConfig != null)
{
break;
}
soln.GetNext();
}
return(tempConfig);
}
public List <List <BinaryOption> > generateRandomVariantsUntilSeconds(VariabilityModel vm, int seconds,
int treshold, int modulu)
{
List <List <BinaryOption> > erglist = new List <List <BinaryOption> >();
var cts = new CancellationTokenSource();
var task = Task.Factory.StartNew(() =>
{
#region task
List <CspTerm> variables = new List <CspTerm>();
Dictionary <BinaryOption, CspTerm> elemToTerm = new Dictionary <BinaryOption, CspTerm>();
Dictionary <CspTerm, BinaryOption> termToElem = new Dictionary <CspTerm, BinaryOption>();
ConstraintSystem S = CSPsolver.getConstraintSystem(out variables, out elemToTerm, out termToElem, vm);
ConstraintSolverSolution soln = S.Solve();
int mod = 0;
while (soln.HasFoundSolution)
{
if (cts.IsCancellationRequested)
{
return(erglist);
}
mod++;
if (mod % modulu != 0)
{
soln.GetNext();
continue;
}
List <BinaryOption> tempConfig = new List <BinaryOption>();
foreach (CspTerm cT in variables)
{
if (soln.GetIntegerValue(cT) == 1)
{
tempConfig.Add(termToElem[cT]);
}
}
if (tempConfig.Contains(null))
{
tempConfig.Remove(null);
}
erglist.Add(tempConfig);
if (erglist.Count == treshold)
{
break;
}
soln.GetNext();
}
return(erglist);
#endregion
}, cts.Token);
if (Task.WaitAny(new[] { task }, TimeSpan.FromMilliseconds(seconds * 1000)) < 0)
{
Console.WriteLine("configsize: " + erglist.Count);
cts.Cancel();
return(erglist);
}
return(erglist);
}
