/// <summary>EP message to <c>sample</c>.</summary>
/// <param name="choice">Incoming message from <c>choice</c>.</param>
/// <param name="probTrue0">Constant value for <c>probTrue0</c>.</param>
/// <param name="probTrue1">Constant value for <c>probTrue1</c>.</param>
/// <returns>The outgoing EP message to the <c>sample</c> argument.</returns>
/// <remarks>
/// <para>The outgoing message is a distribution matching the moments of <c>sample</c> as the random arguments are varied. The formula is <c>proj[p(sample) sum_(choice) p(choice) factor(sample,choice,probTrue0,probTrue1)]/p(sample)</c>.</para>
/// </remarks>
public static Bernoulli SampleAverageConditional(Bernoulli choice, double probTrue0, double probTrue1)
{
Bernoulli result = new Bernoulli();
if (choice.IsPointMass)
{
return(SampleConditional(choice.Point, probTrue0, probTrue1));
}
#if FAST
result.SetProbTrue(choice.GetProbFalse() * probTrue0 + choice.GetProbTrue() * probTrue1);
#else
// This method is more numerically stable but slower.
// let oX = log(p(X)/(1-p(X))
// let oY = log(p(Y)/(1-p(Y))
// oX = log( (TT*sigma(oY) + TF*sigma(-oY))/(FT*sigma(oY) + FF*sigma(-oY)) )
// = log( (TT*exp(oY) + TF)/(FT*exp(oY) + FF) )
// = log( (exp(oY) + TF/TT)/(exp(oY) + FF/FT) ) + log(TT/FT)
// ay = log(TF/TT)
// by = log(FF/FT)
// offset = log(TT/FT)
if (probTrue0 == 0 || probTrue1 == 0)
{
throw new ArgumentException("probTrue is zero");
}
double ay = Math.Log(probTrue0 / probTrue1);
double by = Math.Log((1 - probTrue0) / (1 - probTrue1));
double offset = MMath.Logit(probTrue1);
result.LogOdds = MMath.DiffLogSumExp(choice.LogOdds, ay, by) + offset;
#endif
return(result);
}
public Binomial(int trialCount, double probSuccess)
{
TrialCount = trialCount;
LogOdds = MMath.Logit(probSuccess);
A = 1;
B = 1;
}
/// <summary>
/// Sets the probability of the binary variable being true
/// </summary>
public void SetProbTrue(double probTrue)
{
if (probTrue < 0 || probTrue > 1)
{
throw new ArgumentOutOfRangeException("probTrue", String.Format("probTrue = {0} is not in [0,1]", probTrue));
}
LogOdds = MMath.Logit(probTrue);
}
/// <summary>
/// Creates a Bernoulli distribution with given probability of being true.
/// </summary>
/// <param name="probTrue">p(true)</param>
public Bernoulli(double probTrue)
{
if (probTrue < 0 || probTrue > 1)
{
throw new ArgumentOutOfRangeException(nameof(probTrue), $"{nameof(probTrue)} = {probTrue} is not in [0,1]");
}
LogOdds = MMath.Logit(probTrue);
}
/// <summary>
/// Sets the probability of the binary variable being false
/// </summary>
public void SetProbFalse(double probFalse)
{
if (probFalse < 0 || probFalse > 1)
{
throw new ArgumentOutOfRangeException(nameof(probFalse), $"{nameof(probFalse)} = {probFalse} is not in [0,1]");
}
LogOdds = -MMath.Logit(probFalse);
}
/// <summary>
/// Creates a Bernoulli distribution with given probability of being true.
/// </summary>
/// <param name="probTrue">p(true)</param>
public Bernoulli(double probTrue)
{
if (probTrue < 0 || probTrue > 1)
{
throw new ArgumentException(String.Format("probTrue = {0} is not in [0,1]", probTrue), "probTrue");
}
LogOdds = MMath.Logit(probTrue);
}
/// <summary>
/// Sets the probability of the binary variable being false
/// </summary>
public void SetProbFalse(double probFalse)
{
if (probFalse < 0 || probFalse > 1)
{
throw new ArgumentException(String.Format("probFalse = {0} is not in [0,1]", probFalse), "probFalse");
}
LogOdds = -MMath.Logit(probFalse);
}
/// <summary>
/// Sets the probability of the binary variable being true
/// </summary>
public void SetProbTrue(int index, double probTrue)
{
if (probTrue < 0 || probTrue > 1)
{
throw new ArgumentException(String.Format("probTrue = {0} is not in [0,1]", probTrue), "probTrue");
}
LogOddsVector[index] = MMath.Logit(probTrue);
}
/// <summary>
/// Evidence message for EP
/// </summary>
/// <param name="logistic">Constant value for 'logistic'.</param>
/// <param name="x">Incoming message from 'x'.</param>
/// <returns>Logarithm of the factor's average value across the given argument distributions</returns>
/// <remarks><para>
/// The formula for the result is <c>log(sum_(x) p(x) factor(logistic,x))</c>.
/// </para></remarks>
public static double LogAverageFactor(double logistic, Gaussian x)
{
if (logistic >= 1.0 || logistic <= 0.0)
{
return(x.GetLogProb(MMath.Logit(logistic)));
}
// p(y,x) = delta(y - 1/(1+exp(-x))) N(x;mx,vx)
// x = log(y/(1-y))
// dx = 1/(y*(1-y))
return(x.GetLogProb(MMath.Logit(logistic)) / (logistic * (1 - logistic)));
}
/// <summary>VMP message to <c>sample</c>.</summary>
/// <param name="choice">Incoming message from <c>choice</c>.</param>
/// <param name="probTrue0">Constant value for <c>probTrue0</c>.</param>
/// <param name="probTrue1">Constant value for <c>probTrue1</c>.</param>
/// <returns>The outgoing VMP message to the <c>sample</c> argument.</returns>
/// <remarks>
/// <para>The outgoing message is the exponential of the average log-factor value, where the average is over all arguments except <c>sample</c>. The formula is <c>exp(sum_(choice) p(choice) log(factor(sample,choice,probTrue0,probTrue1)))</c>.</para>
/// </remarks>
public static Bernoulli SampleAverageLogarithm(Bernoulli choice, double probTrue0, double probTrue1)
{
Bernoulli result = new Bernoulli();
if (choice.IsPointMass)
{
return(SampleConditional(choice.Point, probTrue0, probTrue1));
}
// log(p(X=true)/p(X=false)) = sum_k p(Y=k) log(ProbTrue[k]/(1-ProbTrue[k]))
result.LogOdds = choice.GetProbFalse() * MMath.Logit(probTrue0) + choice.GetProbTrue() * MMath.Logit(probTrue1);
return(result);
}
/// <summary>
/// Used to compute log odds in the above operator
/// </summary>
/// <param name="trueCount"></param>
/// <param name="falseCount"></param>
/// <returns></returns>
internal static double ComputeLogOdds(double trueCount, double falseCount)
{
if (falseCount == Double.PositiveInfinity)
{
// compute log odds from prob true
return(MMath.Logit(trueCount));
}
else if ((trueCount == 0) || (falseCount == 0))
{
throw new ImproperMessageException(new Beta(trueCount, falseCount));
}
return(MMath.Digamma(trueCount) - MMath.Digamma(falseCount));
}
/// <summary>
/// VMP message to 'sample'.
/// </summary>
/// <param name="index">Incoming message from 'index'.</param>
/// <param name="ProbTrue">Constant value for 'probTrue'.</param>
/// <returns>The outgoing VMP message to the 'sample' argument.</returns>
/// <remarks><para>
/// The outgoing message is the exponential of the integral of the log-factor times incoming messages, over all arguments except 'sample'.
/// The formula is <c>int log(f(sample,x)) q(x) dx</c> where <c>x = (index,probTrue)</c>.
/// </para></remarks>
public static Bernoulli SampleAverageLogarithm(Discrete index, double[] ProbTrue)
{
Bernoulli result = new Bernoulli();
// E[sum_k I(Y=k) (X*log(ProbTrue[k]) + (1-X)*log(1-ProbTrue[k]))]
// = X*(sum_k p(Y=k) log(ProbTrue[k])) + (1-X)*(sum_k p(Y=k) log(1-ProbTrue[k]))
// p(X=true) =propto prod_k ProbTrue[k]^p(Y=k)
// log(p(X=true)/p(X=false)) = sum_k p(Y=k) log(ProbTrue[k]/(1-ProbTrue[k]))
double s = 0;
for (int i = 0; i < index.Dimension; i++)
{
s += index[i] * MMath.Logit(ProbTrue[i]);
}
result.LogOdds = s;
return(result);
}
[NoTriggers] // see VmpTests.AdditiveSparseBlockmodel
public static double A([Proper, SkipIfUniform] Gaussian x, double a)
{
double m, v;
x.GetMeanAndVariance(out m, out v);
double newa = .5 - (MMath.Logit(a) - m) / v;
double normala = MMath.Logistic(m + (1.0 - 2.0 * a) * v / 2.0);
if (logSumExpBound(m, v, normala) > logSumExpBound(m, v, newa))
{
return(newa);
}
else
{
return(normala);
}
}
/// <summary>EP message to <c>choice</c>.</summary>
/// <param name="sample">Incoming message from <c>sample</c>. Must be a proper distribution. If uniform, the result will be uniform.</param>
/// <param name="probTrue0">Constant value for <c>probTrue0</c>.</param>
/// <param name="probTrue1">Constant value for <c>probTrue1</c>.</param>
/// <returns>The outgoing EP message to the <c>choice</c> argument.</returns>
/// <remarks>
/// <para>The outgoing message is a distribution matching the moments of <c>choice</c> as the random arguments are varied. The formula is <c>proj[p(choice) sum_(sample) p(sample) factor(sample,choice,probTrue0,probTrue1)]/p(choice)</c>.</para>
/// </remarks>
/// <exception cref="ImproperMessageException">
/// <paramref name="sample" /> is not a proper distribution.</exception>
public static Bernoulli ChoiceAverageConditional([SkipIfUniform] Bernoulli sample, double probTrue0, double probTrue1)
{
Bernoulli result = new Bernoulli();
if (sample.IsPointMass)
{
return(ChoiceConditional(sample.Point, probTrue0, probTrue1));
}
#if FAST
double p1 = sample.GetProbFalse() * (1 - probTrue1) + sample.GetProbTrue() * probTrue1;
double p0 = sample.GetProbFalse() * (1 - probTrue0) + sample.GetProbTrue() * probTrue0;
double sum = p0 + p1;
if (sum == 0.0)
{
throw new AllZeroException();
}
else
{
result.SetProbTrue(p1 / sum);
}
#else
// This method is more numerically stable but slower.
// ax = log(FT/TT)
// bx = log(FF/TF)
// offset = log(TT/TF)
if (probTrue0 == 0 || probTrue1 == 0)
{
throw new ArgumentException("probTrue is zero");
}
double ax = -MMath.Logit(probTrue1);
double bx = -MMath.Logit(probTrue0);
double offset = Math.Log(probTrue1 / probTrue0);
result.LogOdds = MMath.DiffLogSumExp(sample.LogOdds, ax, bx) + offset;
#endif
return(result);
}
/// <summary>
/// Creates a list of Bernoullis of the specified size, each set to have the specified probability of being true.
/// </summary>
/// <param name="size">The size of the list</param>
/// <param name="probTrue">The probability of each Bernoulli being true</param>
public SparseBernoulliList(int size, double probTrue)
{
LogOddsVector = ApproximateSparseVector.Constant(size, MMath.Logit(probTrue), DefaultSparsity);
}
/// <summary>
/// EP message to 'x'
/// </summary>
/// <param name="logistic">Constant value for 'logistic'.</param>
/// <returns>The outgoing EP message to the 'x' argument</returns>
/// <remarks><para>
/// The outgoing message is the factor viewed as a function of 'x' conditioned on the given values.
/// </para></remarks>
public static Gaussian XAverageConditional(double logistic)
{
return(Gaussian.PointMass(MMath.Logit(logistic)));
}
/// <include file='FactorDocs.xml' path='factor_docs/message_op_class[@name="DiscreteAreEqualOp"]/message_doc[@name="AreEqualAverageConditional(int, Discrete)"]/*'/>
public static Bernoulli AreEqualAverageConditional(int A, Discrete B)
{
return(Bernoulli.FromLogOdds(MMath.Logit(B[A])));
}
/// <include file='FactorDocs.xml' path='factor_docs/message_op_class[@name="DiscreteAreEqualOp"]/message_doc[@name="AreEqualAverageConditional(Discrete, Discrete)"]/*'/>
public static Bernoulli AreEqualAverageConditional(Discrete A, Discrete B)
{
return(Bernoulli.FromLogOdds(MMath.Logit(A.ProbEqual(B))));
}
/// <summary>
/// Sets this BernoulliIntegerSubset distribution to the weighted sum of two other such distributions
/// </summary>
/// <param name="weight1"></param>
/// <param name="value1"></param>
/// <param name="weight2"></param>
/// <param name="value2"></param>
/// <remarks>Not yet implemented</remarks>
public void SetToSum(double weight1, BernoulliIntegerSubset value1, double weight2, BernoulliIntegerSubset value2)
{
if (weight1 + weight2 == 0)
{
SetToUniform();
}
else
{
LogOddsVector.SetToFunction(value1.GetProbTrueVector(), value2.GetProbTrueVector(), (x, y) => MMath.Logit((weight1 * x + weight2 * y) / (weight1 + weight2)));
}
}