func NewDir(c *config.Config, fmtr lattice.Formatter, showPr bool, dirname string) (*Dir, error) {
samples := c.OutputFile(dirname)
err := os.MkdirAll(samples, 0775)
if err != nil {
return nil, err
}
var prfmtr lattice.PrFormatter
if showPr {
prfmtr = fmtr.PrFormatter()
}
r := &Dir{
config: c,
fmtr: fmtr,
prfmtr: prfmtr,
dir: samples,
}
return r, nil
}
func NewWalker(conf *config.Config, estimatingWalks int, max bool) (*Walker, error) {
ests, err := conf.BytesFloatMultiMap("uniprox-weight-ests")
if err != nil {
return nil, err
}
prs, err := conf.BytesFloatMultiMap("uniprox-selection-prs")
if err != nil {
return nil, err
}
miner := &Walker{
EstimatingWalks: estimatingWalks,
Ests: ests,
Prs: prs,
Max: max,
}
miner.Walker = *walker.NewWalker(conf, graple.MakeAbsorbingWalk(graple.MakeSample(miner), make(chan error)))
return miner, nil
}
func NewUnique(conf *config.Config, fmtr lattice.Formatter, reporter miners.Reporter, histogramName string) (*Unique, error) {
seen, err := conf.BytesIntMultiMap("unique-seen")
if err != nil {
return nil, err
}
var histogram io.WriteCloser = nil
if histogramName != "" {
histogram, err = os.Create(conf.OutputFile(histogramName + ".csv"))
if err != nil {
return nil, err
}
}
u := &Unique{
fmtr: fmtr,
Seen: seen,
Reporter: reporter,
histogram: histogram,
}
return u, nil
}
func NewFile(c *config.Config, fmtr lattice.Formatter, showPr bool, patternsFilename, embeddingsFilename, namesFilename, matricesFilename, prsFilename string) (*File, error) {
patterns, err := os.Create(c.OutputFile(patternsFilename + fmtr.FileExt()))
if err != nil {
return nil, err
}
embeddings, err := os.Create(c.OutputFile(embeddingsFilename + fmtr.FileExt()))
if err != nil {
return nil, err
}
names, err := os.Create(c.OutputFile(namesFilename))
if err != nil {
return nil, err
}
var matrices io.WriteCloser
var prs io.WriteCloser
var prfmtr lattice.PrFormatter
if showPr {
prfmtr = fmtr.PrFormatter()
if prfmtr != nil {
prs, err = os.Create(c.OutputFile(prsFilename))
if err != nil {
return nil, err
}
matrices, err = os.Create(c.OutputFile(matricesFilename))
if err != nil {
return nil, err
}
}
}
r := &File{
config: c,
fmtr: fmtr,
prfmtr: prfmtr,
patterns: patterns,
embeddings: embeddings,
names: names,
matrices: matrices,
prs: prs,
}
return r, nil
}
func NewItemSets(config *config.Config, min, max int) (i *ItemSets, err error) {
parents, err := config.IntsIntsMultiMap("itemsets-parents")
if err != nil {
return nil, err
}
parentCount, err := config.IntsIntMultiMap("itemsets-parent-count")
if err != nil {
return nil, err
}
children, err := config.IntsIntsMultiMap("itemsets-children")
if err != nil {
return nil, err
}
childCount, err := config.IntsIntMultiMap("itemsets-child-count")
if err != nil {
return nil, err
}
canonKids, err := config.IntsIntsMultiMap("itemsets-canon-kids")
if err != nil {
return nil, err
}
canonKidCount, err := config.IntsIntMultiMap("itemsets-canon-kid-count")
if err != nil {
return nil, err
}
embeddings, err := config.IntsIntsMultiMap("itemsets-embeddings")
if err != nil {
return nil, err
}
i = &ItemSets{
MinItems: min,
MaxItems: max,
Parents: parents,
ParentCount: parentCount,
Children: children,
ChildCount: childCount,
CanonKids: canonKids,
CanonKidCount: canonKidCount,
Embeddings: embeddings,
config: config,
}
return i, nil
}
func NewDigraph(config *config.Config, dc *Config) (g *Digraph, err error) {
if dc.MaxEdges <= 0 {
dc.MaxEdges = int(math.MaxInt32)
}
if dc.MaxVertices <= 0 {
dc.MaxVertices = int(math.MaxInt32)
}
if dc.MinEdges > dc.MaxEdges {
dc.MinEdges = dc.MaxEdges - 1
}
if dc.MinVertices > dc.MaxVertices {
dc.MinVertices = dc.MaxVertices - 1
}
nodeAttrs, err := config.IntJsonMultiMap("digraph-node-attrs")
if err != nil {
return nil, err
}
parents, err := config.MultiMap("digraph-parents")
if err != nil {
return nil, err
}
parentCount, err := config.BytesIntMultiMap("digraph-parent-count")
if err != nil {
return nil, err
}
children, err := config.MultiMap("digraph-children")
if err != nil {
return nil, err
}
childCount, err := config.BytesIntMultiMap("digraph-child-count")
if err != nil {
return nil, err
}
canonKids, err := config.MultiMap("digraph-canon-kids")
if err != nil {
return nil, err
}
canonKidCount, err := config.BytesIntMultiMap("digraph-canon-kid-count")
if err != nil {
return nil, err
}
embeddings, err := config.SubgraphEmbeddingMultiMap("digraph-embeddings")
if err != nil {
return nil, err
}
var unsupEmbs subgraph_embedding.MultiMap = nil
if dc.Mode&EmbeddingPruning == EmbeddingPruning {
unsupEmbs, err = config.SubgraphEmbeddingMultiMap("digraph-unsupported-embeddings")
if err != nil {
return nil, err
}
}
var overlap subgraph_overlap.MultiMap = nil
if dc.Mode&OverlapPruning == OverlapPruning {
overlap, err = config.SubgraphOverlapMultiMap("digraph-overlap")
if err != nil {
return nil, err
}
}
exts, err := config.BytesExtensionMultiMap("digraph-extensions")
if err != nil {
return nil, err
}
var unexts bytes_extension.MultiMap
if dc.Mode&ExtensionPruning == ExtensionPruning {
unexts, err = config.BytesExtensionMultiMap("digraph-unsupported-extensions")
if err != nil {
return nil, err
}
}
frequency, err := config.BytesIntMultiMap("digraph-pattern-frequency")
if err != nil {
return nil, err
}
g = &Digraph{
Config: *dc,
NodeAttrs: nodeAttrs,
Embeddings: embeddings,
UnsupEmbs: unsupEmbs,
Overlap: overlap,
Extensions: exts,
UnsupExts: unexts,
Parents: parents,
ParentCount: parentCount,
Children: children,
ChildCount: childCount,
CanonKids: canonKids,
CanonKidCount: canonKidCount,
Frequency: frequency,
config: config,
pool: pool.New(config.Workers()),
}
return g, nil
}