v0.21.0

Breaking changes

• Updated Rust Polars to 0.44.2 (#1271).
  • Minimum supported Rust version (MSRV) is now 1.82.0.
  • $reshape()'s nested_type argument is removed.
  • $approx_n_unique() no longer works on Categorical type.
• <Series>$compare() is removed. (#1272)

Deprecations

• Passing a single data.frame to pl$DataFrame() or pl$LazyFrame() to convert a
  data.frame to a polars DataFrame or LazyFrame is deprecated and a warning will
  be shown. Use as_polars_df() or as_polars_lf() instead (#1275).

Bug fixes

• Maintain level order when converting Enums to factors (#1252, @andyquinterom).

Full Changelog: v0.19.0...v0.21.0

lib-v0.44.0

fix: fix csv download issue on windows (#1292)

v0.20.0

Breaking changes

• Updated rust-polars to 0.43.1 (#1230).
• In pl$scan_ipc() and pl$read_ipc(), the argument memory_map is removed
  (#1230).
• In $serialize(), in the field schema, the field inner is renamed fields,
  and the fields output_schema and filter are removed (#1230).

New features

• New method $cast() for DataFrame and LazyFrame (#1219).
• New argument strict in $drop() to determine whether unknown column names
  should trigger an error (#1220).
• New method $to_dummies() for DataFrame (#1225).
• New argument include_file_paths in pl$scan_csv() and pl$read_csv() (#1235).
• New method $join_where() for DataFrame and LazyFrame to perform
  inequality joins (#1237).

Bug fixes

• Converting data of datatype Null to R doesn't error anymore. It now creates
  a column filled with NA (#1217).

New Contributors

• @collioud made their first contribution in #1238

Full Changelog: v0.19.0...v0.20.0

lib-v0.43.0

test: the latest nanoarrow supports utf8view type (#1257)

v0.19.1

This is a maintenance relase. No user facing changes.

lib-v0.42.1

docs: fix some typos in DEVELOPMENT.md (#1211)

v0.19.0

Breaking changes

• Updated rust-polars to unreleased 2024-08-20, after 0.42.0 (#1183).
• $describe_plan() and $describe_optimized_plan() are removed. Use
  respectively $explain(optimized = FALSE) and $explain() instead (#1182).
• The parameter inherit_optimization is removed from all functions that had it
  (#1183).
• In $write_parquet() and $sink_parquet(), the parameter data_pagesize_limit
  is renamed data_page_size (#1183).
• The LazyFrame method $get_optimization_toggle() is removed, and
  $set_optimization_toggle() is renamed $optimization_toggle() (#1183).
• In $unpivot(), the parameter streamable is removed (#1183).
• Some functions have a parameter future that determines the compatibility level
  when exporting Polars' internal data structures. This parameter is renamed
  compat_level, which takes FALSE for the oldest flavor (more compatible)
  and TRUE for the newest one (less compatible). It can also take an integer
  determining a specific compatibility level when more are added in the future.
  For now, future = FALSE can be replaced by compat_level = FALSE (#1183).
• In $scan_parquet() and $read_parquet(), the default value of
  hive_partitioning is now NULL (#1189).
• In $dt$epoch(), the argument tu is renamed to time_unit (#1196).
• In $fill_nan() for DataFrame, LazyFrame and Expr, the argument is
  renamed value (#1198).
• $shift_and_fill() is removed and replaced by a new argument fill_value in
  $shift(). $shift_and_fill(fill_value, periods) can be replaced by
  $shift(n, fill_value) (#1201).
• In $shift() for various Expr, the argument periods is renamed n (#1201).
• In $clip(), arguments min and max are renamed lower_bound and
  upper_bound (#1203).
• $clip_min() and $clip_max() are removed. Use $clip() with only
  lower_bound or upper_bound instead (#1203).
• In $write_csv and $sink_csv(), the argument quote is renamed
  quote_char (#1206).

New features

• New method $str$extract_many() (#1163).
• Converting a nanoarrow_array with zero rows to an RPolarsDataFrame via
  as_polars_df() now keeps the original schema (#1177).
• $write_parquet() has two new arguments partition_by and
  partition_chunk_size_bytes to write a DataFrame to a hive-partitioned
  directory (#1183).
• New method $bin$size() (#1183).
• In $scan_parquet() and $read_parquet(), the parallel argument can take
  the new value "prefiltered" (#1183).
• $scan_parquet(), $scan_ipc() and $read_parquet() have a new argument
  include_file_paths to automatically add a column containing the path to the
  source file(s) (#1183).
• $scan_ipc() can read a hive-partitioned directory with its new arguments
  hive_partitioning, hive_schema, and try_parse_hive_dates (#1183).
• $scan_parquet() and $read_parquet() gain two new arguments for more control
  on importing hive partitions: hive_schema and try_parse_hive_dates (#1189).
• New method $gather_every() for LazyFrame and DataFrame (#1199).
• $glimpse() for DataFrame has two new arguments max_items_per_column and
  max_colname_length (#1200).
• New method $list$sample() (#1204).
• New argument coalesce in $join_asof() (#1205).
• New argument maintain_order in $list$unique() (#1207).

Other changes

• In $unnest() for DataFrame and LazyFrame, the names argument is removed
  and replaced by .... This doesn't change the previous behavior, e.g.
  df$unnest(names = c("a", "b")) still works (#1170).

---

Full Changelog: v0.18.0...v0.19.0

lib-v0.42.0

chore: bump serde_json from 1.0.125 to 1.0.127 in /src/rust (#1209)

Signed-off-by: dependabot[bot] <[email protected]>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

v0.18.0

Breaking changes

• Updated rust-polars to 0.41.3 (#1147, #1156).
• In $n_chunks(), the default value of strategy now is "first" (#1137).
• $sample() for Expr and DataFrame (#1136):
  • the argument frac is renamed fraction;
  • all the arguments except n must be named;
  • for the Expr method only, the first argument is now n (it was already the
    case for the DataFrame method);
  • for the Expr method only, the default value for with_replacement is now
    FALSE (it was already the case for the DataFrame method).
• $melt() had several changes (#1147):
  • melt() is renamed $unpivot().
  • Some arguments were renamed: id_vars is now index, value_vars is now
    on.
  • The order of arguments has changed: on is now first, then index. The
    order of the other arguments hasn't changed. Note that on can be unnamed
    but all the other arguments must be named.
• pivot() had several changes (#1147):
  • The argument columns is renamed on.
  • The order of arguments has changed: on is now first, then index and
    values. The order of the other arguments hasn't changed. Note that on
    can be unnamed but all the other arguments must be named.
• In $write_parquet() and $sink_parquet(), the default value of argument
  statistics is now TRUE and can take other values than TRUE/FALSE (#1147).
• In $dt$truncate() and $dt$round(), the argument offset has been removed.
  Use $dt$offset_by() after those functions instead (#1147).
• In $top_k() and $bottom_k() for Expr, the arguments nulls_last,
  maintain_order and multithreaded have been removed. If any null values
  are in the top/bottom k values, they will always be positioned last (#1147).
• $replace() has been split in two functions depending on the desired
  behaviour (#1147):
  • $replace() recodes some values in the column, leaving all other values
    unchanged. Compared to the previous version, it doesn't use the arguments
    default and return_dtype anymore.
  • $replace_strict() replaces all values by different values. If a value
    doesn't have a specific mapping, it is replaced by the default value.
• $str$concat() is deprecated, use $str$join() (with the same arguments)
  instead (#1147).
• In pl$date_range() and pl$date_ranges(), the arguments time_unit and
  time_zone have been removed. They were deprecated in previous versions
  (#1147).
• In $join(), when how = "cross", on, left_on and right_on must be
  NULL (#1147).

New features

• New method $has_nulls() (#1133).
• New method $list$explode() (#1139).
• $over() gains a new argument order_by to specify the order of values
  within each group. This is useful when the operation depends on the order of
  values, such as $shift() (#1147).
• $value_counts() gains an argument normalize to give relative frequencies
  of unique values instead of their count (#1147).

New Contributors

• @ju6ge made their first contribution in #1135
• @shikokuchuo made their first contribution in #1160

Full Changelog: v0.17.0...v0.18.0

lib-v0.41.0

test: tempolary disable the test of `pl$mem_address` (#1161)