Python 0.6.2

Bugfixes

• Meatdata.schema was returning a modified schema, this is fixed to return a copy of
  the original schema instead (@benjeffery, #3129, #3130)

Python 0.6.1

Bugfixes

• Fix to TreeSequence.pair_coalescence_counts output dimension when
  provided with time windows containing no nodes (@nspope,
  #3046, #3058)

• Fix to TreeSequence.pair_coalescence_counts to normalise by non-missing
  span if span_normalise=True. This resolves a bug where
  TreeSequence.pair_coalescence_rates would return incorrect values for
  intervals with missing trees. (@natep, #3053, #3059)

• Fix to TreeSequence.pair_coalescence_rates causing an
  assertion to be triggered by floating point error, when all coalescence events are inside a single time window (@natep, #3035, #3038)

Features

• Add support for fixed-length arrays in metadata struct codec using the length property.
  (@benjeffery, #3088,#3090)

• Add a new TreeSequence.pca method that uses randomized linear algebra
  to find the top eigenvectors/values of the genetic relatedness matrix
  (@hanbin973, @petrelharp, #3008)

• Add methods on TreeSequence to efficiently get table metadata as a
  numpy structured array. (@benjeffery, #3098)

• Add Python 3.13 support (@benjeffery, #3107)

• Add a preamble argument to draw_svg() methods to allow adding arbitrary extra
  graphics (e.g. legends) to SVG plots (@hyanwong, issue:3086`, #3121)

C API C_1.1.4

Changes

• Added the TSK_TRACE_ERRORS macro to enable tracing of errors in the C library.
  This is useful for debugging as errors will print to stderr when set.
  (@jeromekelleher, #3095).

Python 0.6.0

Breaking Changes

• The definition of TreeSequence.genetic_relatedness and
  TreeSequence.genetic_relatedness_weighted are changed
  to average over sample sets, rather than summing over them.
  For computation with diploid sample sets, this will change the result
  by a factor of four; for larger sample sets it will now produce
  sensible values that are comparable between sample sets of different sizes.
  The default for these methods is also changed to polarised=True,
  but the output is unchanged for centre=True (the default).
  See the documentation for these methods for more discussion.
  (@petrelharp, @mmosmond, #1623)

Bugfixes

• Fix to TreeSequence.genetic_relatedness with indexes=None and
  proportion=True. (@petrelharp, #2984, #1623)

• Fix to TreeSequence.general_stat when using non-strict summary functions
  in the presence of non-ancestral material (very rare).
  (@petrelharp, #2983, #1623)

• Printing tskit.MetadataSchema(schema=None) now shows "Null_schema" rather
  than None, to avoid confusion (@hyanwong, #2720)

• Limit output HTML when a tree sequence is displayed that has a large amount of metadata.
  (@benjeffery, #2999)

• Fix warning in draw_svg to use correct warnings module.
  (@duncanMR, #2870, #2871)

Features

• Add the centre option to TreeSequence.genetic_relatedness and
  TreeSequence.genetic_relatedness_weighted.
  (@petrelharp, @mmosmond, #1623)

• Edges now have an .interval attribute returning a tskit.Interval object.
  (@hyanwong, #2531)

• Variants now have a states() method that returns the genotypes as an
  (inefficient) array of strings, rather than integer indexes, to
  aid comparison of genetic variation (@hyanwong, #2617)

• Added distance_between that calculates the total distance between two nodes in a tree.
  (@Billyzhang1229, #2771)

• Added genetic_relatedness_matrix method to compute
  pairwise genetic relatedness between sample sets.
  (@jeromekelleher, @petrelharp, #2823)

• Add TreeSequence.extend_haplotypes method that extends ancestral haplotypes
  using recombination information, leading to unary nodes in many trees and
  fewer edges. (@petrelharp, @hfr1tz3, :user: nspope,
  @avabamf, #2651, #2938)

• Add Table.drop_metadata to make clearing metadata from tables easy.
  (@jeromekelleher, #2944)

• Add Interval.mid and Tree.mid properties to return the midpoint of the interval.
  (@currocam, #2960)

• Added genetic_relatedness_vector method to compute product of genetic relatedness
  matrix and weight vector.
  (@petrelharp, #2980)

• Added pair_coalescence_counts method to calculate coalescence events per node or time
  interval, pair_coalescence_quantiles method to estimate quantiles of pair
  coalescence times using empirical CDF inversion, and pair_coalescence_rates method to
  estimate instantaneous rates of pair coalescence within time intervals from the empirical CDF.
  (@nspope, #2915, #2976, #2985)

• Add provenance information to the HTML notebook representation of a tree sequence.
  (@benjeffery, #3001)

• The .draw_svg() methods can add annotated genomic regions (e.g. genes) to the
  x-axis. (@hyanwong, #3002)

• Added a node_titles and a mutation_titles parameter to .draw_svg() methods
  which assigns a string to node and mutation symbols, commonly shown on mouseover. This
  can reduce label clutter while retaining useful info (@hyanwong, #3007)

• Added (currently undocumented) use of the order parameter in Tree.draw_svg() to
  pass a subset of nodes, so subtrees can be visually collapsed. Additionally, an option
  pack_untracked_polytomies allows large polytomies involving untracked samples to
  be summarised as a dotted line (@hyanwong, #3011 #3010, #3012)

• Added a title parameter to .draw_svg() methods (@hyanwong, #3015)

• Add comma separation to all display numbers. (@benjeffery, #3017, #3018)

• Added Tree.ancestors(u) method. (@hyanwong, #2706, #3021)

• Add resources section to provenance schema. (@benjeffery, #3016)

• Add Tree.rf_distance method to calculate the unweighted Robinson-Foulds distance
  between two trees. (@Billyzhang1229, #995, #2643, #3032)

C API C_1.1.3

Features

• Add the tsk_treeseq_extend_haplotypes method that can compress a tree sequence
  by extending edges into adjacent trees and thus creating unary nodes in those
  trees (@petrelharp, @hfr1tze, @avabamf, #2651, #2938).

Python 0.5.8

• Add support for numpy 2 (@jeromekelleher, @benjeffery, #2964)

Python 0.5.7

Breaking Changes

• The VCF writing methods (ts.write_vcf, ts.as_vcf) now error if a site with
  position zero is encountered. The VCF spec does not allow zero position sites.
  Suppress this error with the allow_position_zero argument.
  (@benjeffery, #2901, #2838)

Bugfixes

• Fix to the folded, expected allele frequency spectrum (i.e.,
  TreeSequence.allele_frequency_spectrum(mode="branch", polarised=False),
  which was half as big as it should have been. (@petrelharp,
  @nspope, #2933)

Python 0.5.6

Breaking Changes

• tskit now requires Python 3.8, as Python 3.7 became end-of-life on 2023-06-27

Features

• Tree.trmca now accepts >2 nodes and returns nicer errors
  (@hyanwong, :pr:2808, #2801, #2070, #2611)

• Add TreeSequence.genetic_relatedness_weighted stats method.
  (@petrelharp, @brieuclehmann, @jeromekelleher,
  #2785, #1246)

• Add TreeSequence.impute_unknown_mutations_time method to return an
  array of mutation times based on the times of associated nodes
  (@duncanMR, #2760, #2758)

• Add asdict to all dataclasses. These are returned when you access a row or
  other tree sequence object. (@benjeffery, #2759, #2719)

Bugfixes

• Fix incompatibility with jsonschema>4.18.6 which caused
  AttributeError: module jsonschema has no attribute _validators
  (@benjeffery, #2844, #2840)

Python 0.5.5

Performance improvements

• Methods like ts.at() which seek to a specified position on the sequence from
  a new Tree instance are now much faster (@molpopgen, #2661).

Features

• Add __repr__ for variants to return a string representation of the raw data
  without spewing megabytes of text (@chriscrsmith, #2695, #2694)

• Add keep_rows method to table classes to support efficient in-place
  table subsetting (@jeromekelleher, #2700)

Bugfixes

• Fix UnicodeDecodeError when calling Variant.alleles on the emscripten platform.
  (@benjeffery, #2754, #2737)

C API C_1.1.2

Performance improvements

• tsk_tree_seek is now much faster at seeking to arbitrary points along
  the sequence from the null tree (@molpopgen, #2661).

Features

• The struct tsk_treeseq_t now has the variables min_time and max_time,
  which are the minimum and maximum among the node times and mutation times,
  respectively. min_time and max_time can be accessed using the functions
  tsk_treeseq_get_min_time and tsk_treeseq_get_max_time, respectively.
  (@szhan, #2612, #2271)

• Add the TSK_SIMPLIFY_NO_FILTER_NODES option to simplify to allow unreferenced
  nodes be kept in the output (@jeromekelleher, @hyanwong,
  #2606, #2619).

• Add the TSK_SIMPLIFY_NO_UPDATE_SAMPLE_FLAGS option to simplify which ensures
  no node sample flags are changed to allow calling code to manage sample status.
  (@jeromekelleher, #2662, #2663).

• Guarantee that unfiltered tables are not written to unnecessarily
  during simplify (@jeromekelleher #2619).

• Add x_table_keep_rows methods to provide efficient in-place table subsetting
  (@jeromekelleher, #2700).

• Add tsk_tree_seek_index function