Package cc.redberry.core.tensor

Examples of cc.redberry.core.tensor.Tensor

53
54
55
56
57
58
59
60
61
62
63
    }

    @Override
    public Tensor toTensor() {
        SimpleTensor[] vars = new SimpleTensor[content.length - 1];
        Tensor temp;
        for (int i = 1; i < content.length; ++i) {
            temp = content[i].toTensor();
            if (!(temp instanceof SimpleTensor) && !(temp instanceof TensorField))
                throw new IllegalArgumentException("Derivative with respect to non simple argument: " + temp);
            vars[i - 1] = (SimpleTensor) temp;
View Full Code Here
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
            Derivative derivative = (Derivative) tensor;

            //Collecting all indices from derivative vars
            TensorIterator it = derivative.iterator();
            IntArrayList usedIndices = new IntArrayList();
            Tensor current;
            while (it.hasNext()) {
                current = it.next();
                if (Derivative.onVarsIndicator.is(it))
                    usedIndices.addAll(TensorUtils.getAllIndicesNames(current));
            }

            //Creating int set from collected indices
            int[] usedIndicesArray = MathUtils.getSortedDistinct(usedIndices.toArray());

            Tensor target = derivative.getTarget();

            //Calculating used indices for AIM transformation
            int[] targetIndices = IndicesUtils.getSortedDistinctIndicesNames(target.getIndices().getFreeIndices());
            int[] currentUsedIndices = MathUtils.intSetDifference(targetIndices, usedIndices.toArray());
            //Applying AIM transformation
            IndexMappingImpl indexMapping = new IndexMappingImpl(currentUsedIndices);
            ApplyIndexMappingTransformation.INSTANCE.perform(target, indexMapping);
        }
View Full Code Here
118
119
120
121
122
123
124
125
126
        for (int i = 0; i < indicesNames.length; ++i)
            newIndicesNames[i] = indicesNames[permutation[i]];

        //processing new tensor
        IndexMappingImpl im = new IndexMappingImpl(new int[0], indicesNames, newIndicesNames);
        Tensor current = tensor.clone();
        ApplyIndexMappingTransformation.INSTANCE.perform(current, im);
        result.add(current);
    }
View Full Code Here
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
            LinkedList<Tensor> tCombination = new LinkedList<>();
            u = 0;
            l = 0;
            for (i = 0; i < combination.length; ++i)
                for (int j = 0; j < combination[i]; ++j) {
                    Tensor temp = samples[i].clone();

                    IndexMappingDirect im = new IndexMappingDirect();
                    IntArray termLow = temp.getIndices().getLower();
                    im.add(termLow, Arrays.copyOfRange(lowArray, l, l + termLow.length()));
                    l += termLow.length();

                    IntArray termUp = temp.getIndices().getUpper();
                    im.add(termUp, Arrays.copyOfRange(upArray, u, u + termUp.length()));
                    u += termUp.length();

                    temp = ApplyIndexMappingDirectTransformation.INSTANCE.perform(temp, im);
                    tCombination.add(temp);
                }

            //creating term & processing combinatorics
            Tensor coefficient;
            if (symmetries == null) {
                Symmetrize symmetrize = new Symmetrize(indices,
                        Symmetries.getFullSymmetriesForSortedIndices(totalUpCount, totalLowCount), false);
                Tensor terms = symmetrize.transform(new Product(tCombination));
                if (terms instanceof Sum)
                    for (Tensor t : terms) {
                        result.add(new Product(coefficient = scalarTensorGenerator.next(), t));
                        coefficients.add(coefficient.clone());
                    }
                else {
                    result.add(new Product(coefficient = scalarTensorGenerator.next(), terms));
                    coefficients.add(coefficient.clone());
                }
            } else {
                Symmetrize symmetrize = new Symmetrize(indices,
                        symmetries, true);
                Tensor terms = symmetrize.transform(new Product(tCombination));
                result.add(new Product(coefficient = scalarTensorGenerator.next(), terms));
                coefficients.add(coefficient.clone());
            }
        }
    }
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
    @Override
    public Tensor transform(Tensor tensor) {
        if (!(tensor instanceof Derivative))
            return tensor;
        Derivative derivative = (Derivative) tensor;
        Tensor target = derivative.getTarget().clone();
        for (int i = 0; i < derivative.getDerivativeOrder(); ++i) {
            target = getDerivative(target, derivative.getVariation(i));
            if (target == null)
                return TensorNumber.createZERO();
        }
View Full Code Here
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
    private Tensor getDerivative(Tensor target, SimpleTensor var) {
        if (target instanceof Sum) {
            Sum sum = (Sum) target;
            TensorIterator it = sum.iterator();
            Tensor current, derivative;
            Sum res = new Sum();
            while (it.hasNext()) {
                current = it.next();
                derivative = getDerivative(current, var);
                if (derivative == null)
                    continue;// it.remove();
                else
                    res.add(derivative);//if (derivative != current)
                //it.set(derivative);
            }
            if (res.isEmpty())
                return null;
            return res.equivalent();
        } else if (target instanceof Product) {
            Product product = (Product) target;
            Tensor derivative;
            List<Tensor> resultProducts = new ArrayList<>();
            for (int i = 0; i < product.size(); ++i) {
                derivative = getDerivative(product.getElements().get(i), var);
                if (derivative == null)
                    continue;
                Product clone = (Product) product.clone();
                clone.getElements().remove(i);
                if (!isOne(derivative))
                    clone.add(derivative);
                resultProducts.add(clone.equivalent());
            }
            if (resultProducts.isEmpty())
                return null;
            if (resultProducts.size() == 1)
                return resultProducts.get(0);
            return new Sum(resultProducts);
        } else if (target.getClass() == SimpleTensor.class) {
            SimpleTensor sp = (SimpleTensor) target;
            if (sp.getName() != var.getName())
                return null;
            if (sp.getIndices().size() == 0)
                return TensorNumber.createONE();
            Product kroneckers = new Product();
            Indices targetIndices = sp.getIndices();
            Indices varIndices = var.getIndices();
            for (int i = 0; i < sp.getIndices().size(); ++i)
                kroneckers.add(CC.createMetricOrKronecker(targetIndices.get(i), varIndices.get(i)));
            return kroneckers.equivalent();
        } else if (target.getClass() == TensorField.class) {
            TensorField field = (TensorField) target;
            Tensor[] args = field.getArgs();
            for (int i = 0; i < args.length; ++i)
                if (getDerivative(args[i], var) != null)
                    return Derivative.createFromInversed(target,var);
            return null;
        } else if (target instanceof AbstractScalarFunction) {
            AbstractScalarFunction func = (AbstractScalarFunction) target;
            Tensor der = getDerivative(func.getInnerTensor(), var);
            if (der == null)
                return null;
            if (isOne(der))
                return func.derivative();
            return new Product(func.derivative(), der);
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
    @Override
    public PatternSplit split(Tensor tensor) {
        if (tensor instanceof Product) {
            TensorIterator it = tensor.iterator();
            Tensor current;
            Product factored = new Product();
            while (it.hasNext()) {
                current = it.next();
                if (splitPattern.factorOut(current)) {
                    factored.add(current);
View Full Code Here
84
85
86
87
88
89
90
91
92
93
94
95
            wrapper.write();
            return newSum;
        } else if (tensor instanceof Product) {
            Product p = new Product();
            TensorIterator iterator = tensor.iterator();
            Tensor current;
            Tensor temp = null;
            int[] lower = ind.getLower().copy();
            int[] upper = ind.getUpper().copy();
            int[] total = new int[length * 2];
            while (iterator.hasNext()) {
                current = iterator.next();
View Full Code Here
111
112
113
114
115
116
117
118
119
120
121
122

        //renaming indices of toCollect tensor
        new RenameContractedIndices(factorsIndicesNames.toArray()).transform(split.term, split.factoredOut);

        Tensor toCollect = split.term;
        Tensor collectedTensor = term;


        IntArrayList uncontractedIndicesToCollect = TensorUtils.getContractedIndicesNames(split.term, split.factoredOut);
        IntArrayList uncontractedIndicesCollected = TensorUtils.getContractedIndicesNames(term, factors.getElements().get(0));
View Full Code Here
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
    @Override
    public Tensor tensorEquvivalent() {
        if (!initialized)
            return pairProduct(factoredOut, term);
        TensorIterator it = factors.iterator();
        Tensor equivalent;
        Tensor current;
        while (it.hasNext()) {
            current = it.next();
            equivalent = current.equivalent();
            if (current != equivalent)
                it.set(equivalent);
        }
        return pairProduct(Transformations.calculateNumbers(factors), term.equivalent());
    }
View Full Code Here
TOP

Related Classes of cc.redberry.core.tensor.Tensor

  • cc.redberry.core.parser.ParseManager
  • cc.redberry.core.parser.ParseNodeScalarFunction
  • cc.redberry.core.parser.ParseTokenDerivative
  • cc.redberry.core.parser.ParseTokenScalarFunction
  • cc.redberry.core.performance.IndexMappingPerformance
  • cc.redberry.core.performance.kv.NaiveSubstitution
  • cc.redberry.core.performance.kv.OneLoopInput
  • cc.redberry.core.performance.Main
  • cc.redberry.core.performance.PairECBenchmark0
  • cc.redberry.core.performance.ProductBijectionPerformanceTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.