polars-df 0.5.0 → 0.7.0

data/Cargo.toml

@@ -1,5 +1,6 @@
 [workspace]
 members = ["ext/polars"]
+resolver = "2"
 
 [patch.crates-io]
 jsonpath_lib = { git = "https://github.com/ritchie46/jsonpath", rev = "24eaf0b4416edff38a4d1b6b17bc4b9f3f047b4b" }

data/README.md

@@ -20,12 +20,12 @@ This library follows the [Polars Python API](https://pola-rs.github.io/polars/py
 Polars.read_csv("iris.csv")
   .lazy
   .filter(Polars.col("sepal_length") > 5)
-  .groupby("species")
+  .group_by("species")
   .agg(Polars.all.sum)
   .collect
 ```
 
-You can follow [Polars tutorials](https://pola-rs.github.io/polars-book/user-guide/introduction.html) and convert the code to Ruby in many cases. Feel free to open an issue if you run into problems.
+You can follow [Polars tutorials](https://pola-rs.github.io/polars-book/user-guide/) and convert the code to Ruby in many cases. Feel free to open an issue if you run into problems.
 
 ## Reference
 
@@ -260,19 +260,19 @@ df["a"].var
 Group
 
 ```ruby
-df.groupby("a").count
+df.group_by("a").count
 ```
 
 Works with all summary statistics
 
 ```ruby
-df.groupby("a").max
+df.group_by("a").max
 ```
 
 Multiple groups
 
 ```ruby
-df.groupby(["a", "b"]).count
+df.group_by(["a", "b"]).count
 ```
 
 ## Combining Data Frames
@@ -348,7 +348,7 @@ df.to_numo
 You can specify column types when creating a data frame
 
 ```ruby
-Polars::DataFrame.new(data, columns: {"a" => Polars::Int32, "b" => Polars::Float32})
+Polars::DataFrame.new(data, schema: {"a" => Polars::Int32, "b" => Polars::Float32})
 ```
 
 Supported types are:
@@ -357,8 +357,10 @@ Supported types are:
 - float - `Float64`, `Float32`
 - integer - `Int64`, `Int32`, `Int16`, `Int8`
 - unsigned integer - `UInt64`, `UInt32`, `UInt16`, `UInt8`
-- string - `Utf8`, `Categorical`
+- string - `Utf8`, `Binary`, `Categorical`
 - temporal - `Date`, `Datetime`, `Time`, `Duration`
+- nested -  `List`, `Struct`, `Array`
+- other - `Object`, `Null`
 
 Get column types
 
@@ -401,13 +403,13 @@ df.plot("a", "b", type: "pie")
 Group data
 
 ```ruby
-df.groupby("c").plot("a", "b")
+df.group_by("c").plot("a", "b")
 ```
 
 Stacked columns or bars
 
 ```ruby
-df.groupby("c").plot("a", "b", stacked: true)
+df.group_by("c").plot("a", "b", stacked: true)
 ```
 
 ## History

data/ext/polars/Cargo.toml

@@ -1,9 +1,10 @@
 [package]
 name = "polars"
-version = "0.5.0"
+version = "0.7.0"
 license = "MIT"
 authors = ["Andrew Kane <andrew@ankane.org>"]
 edition = "2021"
+rust-version = "1.73.0"
 publish = false
 
 [lib]
@@ -11,22 +12,25 @@ crate-type = ["cdylib"]
 
 [dependencies]
 ahash = "0.8"
-magnus = "0.5"
-polars-core = "0.29.0"
+chrono = "0.4"
+either = "1.8"
+magnus = "0.6"
+polars-core = "=0.35.2"
+polars-parquet = "=0.35.2"
 serde_json = "1"
 smartstring = "1"
 
 [dependencies.polars]
-version = "0.29.0"
+version = "=0.35.2"
 features = [
     "abs",
     "approx_unique",
-    "arange",
     "arg_where",
     "asof_join",
     "avro",
     "binary_encoding",
     "concat_str",
+    "cov",
     "cse",
     "csv",
     "cum_agg",
@@ -37,22 +41,23 @@ features = [
     "diff",
     "dot_product",
     "dtype-full",
-    "dynamic_groupby",
+    "dynamic_group_by",
     "ewma",
     "extract_jsonpath",
     "fmt",
     "horizontal_concat",
     "interpolate",
     "ipc",
-    "is_first",
+    "is_first_distinct",
     "is_in",
+    "is_last_distinct",
     "is_unique",
     "json",
     "lazy",
     "lazy_regex",
     "list_count",
     "list_eval",
-    "list_take",
+    "list_gather",
     "list_to_struct",
     "log",
     "meta",
@@ -62,12 +67,14 @@ features = [
     "parquet",
     "partition_by",
     "pct_change",
+    "peaks",
     "performant",
     "pivot",
     "product",
     "propagate_nans",
     "random",
     "rank",
+    "range",
     "reinterpret",
     "repeat_by",
     "rolling_window",
@@ -77,9 +84,10 @@ features = [
     "semi_anti_join",
     "serde-lazy",
     "sign",
+    "sql",
     "string_encoding",
-    "string_from_radix",
-    "string_justify",
+    "string_pad",
+    "string_to_integer",
     "strings",
     "timezones",
     "to_dummies",

data/ext/polars/src/batched_csv.rs

@@ -1,4 +1,4 @@
-use magnus::{RArray, Value};
+use magnus::{prelude::*, RArray, Value};
 use polars::io::mmap::MmapBytesReader;
 use polars::io::RowCount;
 use polars::prelude::read_impl::OwnedBatchedCsvReader;
@@ -24,31 +24,31 @@ impl RbBatchedCsv {
     pub fn new(arguments: &[Value]) -> RbResult<Self> {
         // start arguments
         // this pattern is needed for more than 16
-        let infer_schema_length: Option<usize> = arguments[0].try_convert()?;
-        let chunk_size: usize = arguments[1].try_convert()?;
-        let has_header: bool = arguments[2].try_convert()?;
-        let ignore_errors: bool = arguments[3].try_convert()?;
-        let n_rows: Option<usize> = arguments[4].try_convert()?;
-        let skip_rows: usize = arguments[5].try_convert()?;
-        let projection: Option<Vec<usize>> = arguments[6].try_convert()?;
-        let sep: String = arguments[7].try_convert()?;
-        let rechunk: bool = arguments[8].try_convert()?;
-        let columns: Option<Vec<String>> = arguments[9].try_convert()?;
-        let encoding: Wrap<CsvEncoding> = arguments[10].try_convert()?;
-        let n_threads: Option<usize> = arguments[11].try_convert()?;
-        let path: PathBuf = arguments[12].try_convert()?;
-        let overwrite_dtype: Option<Vec<(String, Wrap<DataType>)>> = arguments[13].try_convert()?;
+        let infer_schema_length = Option::<usize>::try_convert(arguments[0])?;
+        let chunk_size = usize::try_convert(arguments[1])?;
+        let has_header = bool::try_convert(arguments[2])?;
+        let ignore_errors = bool::try_convert(arguments[3])?;
+        let n_rows = Option::<usize>::try_convert(arguments[4])?;
+        let skip_rows = usize::try_convert(arguments[5])?;
+        let projection = Option::<Vec<usize>>::try_convert(arguments[6])?;
+        let separator = String::try_convert(arguments[7])?;
+        let rechunk = bool::try_convert(arguments[8])?;
+        let columns = Option::<Vec<String>>::try_convert(arguments[9])?;
+        let encoding = Wrap::<CsvEncoding>::try_convert(arguments[10])?;
+        let n_threads = Option::<usize>::try_convert(arguments[11])?;
+        let path = PathBuf::try_convert(arguments[12])?;
+        let overwrite_dtype = Option::<Vec<(String, Wrap<DataType>)>>::try_convert(arguments[13])?;
         // TODO fix
-        let overwrite_dtype_slice: Option<Vec<Wrap<DataType>>> = None; // arguments[14].try_convert()?;
-        let low_memory: bool = arguments[15].try_convert()?;
-        let comment_char: Option<String> = arguments[16].try_convert()?;
-        let quote_char: Option<String> = arguments[17].try_convert()?;
-        let null_values: Option<Wrap<NullValues>> = arguments[18].try_convert()?;
-        let try_parse_dates: bool = arguments[19].try_convert()?;
-        let skip_rows_after_header: usize = arguments[20].try_convert()?;
-        let row_count: Option<(String, IdxSize)> = arguments[21].try_convert()?;
-        let sample_size: usize = arguments[22].try_convert()?;
-        let eol_char: String = arguments[23].try_convert()?;
+        let overwrite_dtype_slice = Option::<Vec<Wrap<DataType>>>::None; // Option::<Vec<Wrap<DataType>>>::try_convert(arguments[14])?;
+        let low_memory = bool::try_convert(arguments[15])?;
+        let comment_char = Option::<String>::try_convert(arguments[16])?;
+        let quote_char = Option::<String>::try_convert(arguments[17])?;
+        let null_values = Option::<Wrap<NullValues>>::try_convert(arguments[18])?;
+        let try_parse_dates = bool::try_convert(arguments[19])?;
+        let skip_rows_after_header = usize::try_convert(arguments[20])?;
+        let row_count = Option::<(String, IdxSize)>::try_convert(arguments[21])?;
+        let sample_size = usize::try_convert(arguments[22])?;
+        let eol_char = String::try_convert(arguments[23])?;
         // end arguments
 
         let null_values = null_values.map(|w| w.0);
@@ -90,7 +90,7 @@ impl RbBatchedCsv {
             .infer_schema(infer_schema_length)
             .has_header(has_header)
             .with_n_rows(n_rows)
-            .with_delimiter(sep.as_bytes()[0])
+            .with_separator(separator.as_bytes()[0])
             .with_skip_rows(skip_rows)
             .with_ignore_errors(ignore_errors)
             .with_projection(projection)