polars-df 0.13.0-x64-mingw-ucrt
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.yardopts +3 -0
- data/CHANGELOG.md +208 -0
- data/Cargo.lock +2556 -0
- data/Cargo.toml +6 -0
- data/LICENSE-THIRD-PARTY.txt +39278 -0
- data/LICENSE.txt +20 -0
- data/README.md +437 -0
- data/lib/polars/3.1/polars.so +0 -0
- data/lib/polars/3.2/polars.so +0 -0
- data/lib/polars/3.3/polars.so +0 -0
- data/lib/polars/array_expr.rb +537 -0
- data/lib/polars/array_name_space.rb +423 -0
- data/lib/polars/batched_csv_reader.rb +104 -0
- data/lib/polars/binary_expr.rb +77 -0
- data/lib/polars/binary_name_space.rb +66 -0
- data/lib/polars/cat_expr.rb +36 -0
- data/lib/polars/cat_name_space.rb +88 -0
- data/lib/polars/config.rb +530 -0
- data/lib/polars/convert.rb +98 -0
- data/lib/polars/data_frame.rb +5191 -0
- data/lib/polars/data_types.rb +466 -0
- data/lib/polars/date_time_expr.rb +1397 -0
- data/lib/polars/date_time_name_space.rb +1287 -0
- data/lib/polars/dynamic_group_by.rb +52 -0
- data/lib/polars/exceptions.rb +38 -0
- data/lib/polars/expr.rb +7256 -0
- data/lib/polars/expr_dispatch.rb +22 -0
- data/lib/polars/functions/aggregation/horizontal.rb +246 -0
- data/lib/polars/functions/aggregation/vertical.rb +282 -0
- data/lib/polars/functions/as_datatype.rb +271 -0
- data/lib/polars/functions/col.rb +47 -0
- data/lib/polars/functions/eager.rb +182 -0
- data/lib/polars/functions/lazy.rb +1329 -0
- data/lib/polars/functions/len.rb +49 -0
- data/lib/polars/functions/lit.rb +35 -0
- data/lib/polars/functions/random.rb +16 -0
- data/lib/polars/functions/range/date_range.rb +136 -0
- data/lib/polars/functions/range/datetime_range.rb +149 -0
- data/lib/polars/functions/range/int_range.rb +51 -0
- data/lib/polars/functions/range/time_range.rb +141 -0
- data/lib/polars/functions/repeat.rb +144 -0
- data/lib/polars/functions/whenthen.rb +96 -0
- data/lib/polars/functions.rb +57 -0
- data/lib/polars/group_by.rb +613 -0
- data/lib/polars/io/avro.rb +24 -0
- data/lib/polars/io/csv.rb +696 -0
- data/lib/polars/io/database.rb +73 -0
- data/lib/polars/io/ipc.rb +275 -0
- data/lib/polars/io/json.rb +29 -0
- data/lib/polars/io/ndjson.rb +80 -0
- data/lib/polars/io/parquet.rb +233 -0
- data/lib/polars/lazy_frame.rb +2708 -0
- data/lib/polars/lazy_group_by.rb +181 -0
- data/lib/polars/list_expr.rb +791 -0
- data/lib/polars/list_name_space.rb +449 -0
- data/lib/polars/meta_expr.rb +222 -0
- data/lib/polars/name_expr.rb +198 -0
- data/lib/polars/plot.rb +109 -0
- data/lib/polars/rolling_group_by.rb +35 -0
- data/lib/polars/series.rb +4444 -0
- data/lib/polars/slice.rb +104 -0
- data/lib/polars/sql_context.rb +194 -0
- data/lib/polars/string_cache.rb +75 -0
- data/lib/polars/string_expr.rb +1495 -0
- data/lib/polars/string_name_space.rb +811 -0
- data/lib/polars/struct_expr.rb +98 -0
- data/lib/polars/struct_name_space.rb +96 -0
- data/lib/polars/testing.rb +507 -0
- data/lib/polars/utils/constants.rb +9 -0
- data/lib/polars/utils/convert.rb +97 -0
- data/lib/polars/utils/parse.rb +89 -0
- data/lib/polars/utils/various.rb +76 -0
- data/lib/polars/utils/wrap.rb +19 -0
- data/lib/polars/utils.rb +130 -0
- data/lib/polars/version.rb +4 -0
- data/lib/polars/whenthen.rb +83 -0
- data/lib/polars-df.rb +1 -0
- data/lib/polars.rb +91 -0
- metadata +138 -0
@@ -0,0 +1,613 @@
|
|
1
|
+
module Polars
|
2
|
+
# Starts a new GroupBy operation.
|
3
|
+
class GroupBy
|
4
|
+
# @private
|
5
|
+
def initialize(df, by, maintain_order: false)
|
6
|
+
@df = df
|
7
|
+
@by = by
|
8
|
+
@maintain_order = maintain_order
|
9
|
+
end
|
10
|
+
|
11
|
+
# Allows iteration over the groups of the group by operation.
|
12
|
+
#
|
13
|
+
# @return [Object]
|
14
|
+
#
|
15
|
+
# @example
|
16
|
+
# df = Polars::DataFrame.new({"foo" => ["a", "a", "b"], "bar" => [1, 2, 3]})
|
17
|
+
# df.group_by("foo", maintain_order: true).each.to_h
|
18
|
+
# # =>
|
19
|
+
# # {"a"=>shape: (2, 2)
|
20
|
+
# # ┌─────┬─────┐
|
21
|
+
# # │ foo ┆ bar │
|
22
|
+
# # │ --- ┆ --- │
|
23
|
+
# # │ str ┆ i64 │
|
24
|
+
# # ╞═════╪═════╡
|
25
|
+
# # │ a ┆ 1 │
|
26
|
+
# # │ a ┆ 2 │
|
27
|
+
# # └─────┴─────┘, "b"=>shape: (1, 2)
|
28
|
+
# # ┌─────┬─────┐
|
29
|
+
# # │ foo ┆ bar │
|
30
|
+
# # │ --- ┆ --- │
|
31
|
+
# # │ str ┆ i64 │
|
32
|
+
# # ╞═════╪═════╡
|
33
|
+
# # │ b ┆ 3 │
|
34
|
+
# # └─────┴─────┘}
|
35
|
+
def each
|
36
|
+
return to_enum(:each) unless block_given?
|
37
|
+
|
38
|
+
temp_col = "__POLARS_GB_GROUP_INDICES"
|
39
|
+
groups_df =
|
40
|
+
@df.lazy
|
41
|
+
.with_row_index(name: temp_col)
|
42
|
+
.group_by(@by, maintain_order: @maintain_order)
|
43
|
+
.agg(Polars.col(temp_col))
|
44
|
+
.collect(no_optimization: true)
|
45
|
+
|
46
|
+
group_names = groups_df.select(Polars.all.exclude(temp_col))
|
47
|
+
|
48
|
+
# When grouping by a single column, group name is a single value
|
49
|
+
# When grouping by multiple columns, group name is a tuple of values
|
50
|
+
if @by.is_a?(::String) || @by.is_a?(Expr)
|
51
|
+
_group_names = group_names.to_series.each
|
52
|
+
else
|
53
|
+
_group_names = group_names.iter_rows
|
54
|
+
end
|
55
|
+
|
56
|
+
_group_indices = groups_df.select(temp_col).to_series
|
57
|
+
_current_index = 0
|
58
|
+
|
59
|
+
while _current_index < _group_indices.length
|
60
|
+
group_name = _group_names.next
|
61
|
+
group_data = @df[_group_indices[_current_index]]
|
62
|
+
_current_index += 1
|
63
|
+
|
64
|
+
yield group_name, group_data
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
68
|
+
# Apply a custom/user-defined function (UDF) over the groups as a sub-DataFrame.
|
69
|
+
#
|
70
|
+
# Implementing logic using a Ruby function is almost always _significantly_
|
71
|
+
# slower and more memory intensive than implementing the same logic using
|
72
|
+
# the native expression API because:
|
73
|
+
|
74
|
+
# - The native expression engine runs in Rust; UDFs run in Ruby.
|
75
|
+
# - Use of Ruby UDFs forces the DataFrame to be materialized in memory.
|
76
|
+
# - Polars-native expressions can be parallelised (UDFs cannot).
|
77
|
+
# - Polars-native expressions can be logically optimised (UDFs cannot).
|
78
|
+
#
|
79
|
+
# Wherever possible you should strongly prefer the native expression API
|
80
|
+
# to achieve the best performance.
|
81
|
+
#
|
82
|
+
# @return [DataFrame]
|
83
|
+
#
|
84
|
+
# @example
|
85
|
+
# df = Polars::DataFrame.new(
|
86
|
+
# {
|
87
|
+
# "id" => [0, 1, 2, 3, 4],
|
88
|
+
# "color" => ["red", "green", "green", "red", "red"],
|
89
|
+
# "shape" => ["square", "triangle", "square", "triangle", "square"]
|
90
|
+
# }
|
91
|
+
# )
|
92
|
+
# df.group_by("color").apply { |group_df| group_df.sample(2) }
|
93
|
+
# # =>
|
94
|
+
# # shape: (4, 3)
|
95
|
+
# # ┌─────┬───────┬──────────┐
|
96
|
+
# # │ id ┆ color ┆ shape │
|
97
|
+
# # │ --- ┆ --- ┆ --- │
|
98
|
+
# # │ i64 ┆ str ┆ str │
|
99
|
+
# # ╞═════╪═══════╪══════════╡
|
100
|
+
# # │ 1 ┆ green ┆ triangle │
|
101
|
+
# # │ 2 ┆ green ┆ square │
|
102
|
+
# # │ 4 ┆ red ┆ square │
|
103
|
+
# # │ 3 ┆ red ┆ triangle │
|
104
|
+
# # └─────┴───────┴──────────┘
|
105
|
+
# def apply(&f)
|
106
|
+
# _dataframe_class._from_rbdf(_df.group_by_apply(by, f))
|
107
|
+
# end
|
108
|
+
|
109
|
+
# Compute aggregations for each group of a group by operation.
|
110
|
+
#
|
111
|
+
# @param aggs [Array]
|
112
|
+
# Aggregations to compute for each group of the group by operation,
|
113
|
+
# specified as positional arguments.
|
114
|
+
# Accepts expression input. Strings are parsed as column names.
|
115
|
+
# @param named_aggs [Hash]
|
116
|
+
# Additional aggregations, specified as keyword arguments.
|
117
|
+
# The resulting columns will be renamed to the keyword used.
|
118
|
+
#
|
119
|
+
# @return [DataFrame]
|
120
|
+
#
|
121
|
+
# @example Compute the aggregation of the columns for each group.
|
122
|
+
# df = Polars::DataFrame.new(
|
123
|
+
# {
|
124
|
+
# "a" => ["a", "b", "a", "b", "c"],
|
125
|
+
# "b" => [1, 2, 1, 3, 3],
|
126
|
+
# "c" => [5, 4, 3, 2, 1]
|
127
|
+
# }
|
128
|
+
# )
|
129
|
+
# df.group_by("a").agg(Polars.col("b"), Polars.col("c"))
|
130
|
+
# # =>
|
131
|
+
# # shape: (3, 3)
|
132
|
+
# # ┌─────┬───────────┬───────────┐
|
133
|
+
# # │ a ┆ b ┆ c │
|
134
|
+
# # │ --- ┆ --- ┆ --- │
|
135
|
+
# # │ str ┆ list[i64] ┆ list[i64] │
|
136
|
+
# # ╞═════╪═══════════╪═══════════╡
|
137
|
+
# # │ a ┆ [1, 1] ┆ [5, 3] │
|
138
|
+
# # │ b ┆ [2, 3] ┆ [4, 2] │
|
139
|
+
# # │ c ┆ [3] ┆ [1] │
|
140
|
+
# # └─────┴───────────┴───────────┘
|
141
|
+
#
|
142
|
+
# @example Compute the sum of a column for each group.
|
143
|
+
# df.group_by("a").agg(Polars.col("b").sum)
|
144
|
+
# # =>
|
145
|
+
# # shape: (3, 2)
|
146
|
+
# # ┌─────┬─────┐
|
147
|
+
# # │ a ┆ b │
|
148
|
+
# # │ --- ┆ --- │
|
149
|
+
# # │ str ┆ i64 │
|
150
|
+
# # ╞═════╪═════╡
|
151
|
+
# # │ a ┆ 2 │
|
152
|
+
# # │ b ┆ 5 │
|
153
|
+
# # │ c ┆ 3 │
|
154
|
+
# # └─────┴─────┘
|
155
|
+
#
|
156
|
+
# @example Compute multiple aggregates at once by passing a list of expressions.
|
157
|
+
# df.group_by("a").agg([Polars.sum("b"), Polars.mean("c")])
|
158
|
+
# # =>
|
159
|
+
# # shape: (3, 3)
|
160
|
+
# # ┌─────┬─────┬─────┐
|
161
|
+
# # │ a ┆ b ┆ c │
|
162
|
+
# # │ --- ┆ --- ┆ --- │
|
163
|
+
# # │ str ┆ i64 ┆ f64 │
|
164
|
+
# # ╞═════╪═════╪═════╡
|
165
|
+
# # │ c ┆ 3 ┆ 1.0 │
|
166
|
+
# # │ a ┆ 2 ┆ 4.0 │
|
167
|
+
# # │ b ┆ 5 ┆ 3.0 │
|
168
|
+
# # └─────┴─────┴─────┘
|
169
|
+
#
|
170
|
+
# @example Or use positional arguments to compute multiple aggregations in the same way.
|
171
|
+
# df.group_by("a").agg(
|
172
|
+
# Polars.sum("b").name.suffix("_sum"),
|
173
|
+
# (Polars.col("c") ** 2).mean.name.suffix("_mean_squared")
|
174
|
+
# )
|
175
|
+
# # =>
|
176
|
+
# # shape: (3, 3)
|
177
|
+
# # ┌─────┬───────┬────────────────┐
|
178
|
+
# # │ a ┆ b_sum ┆ c_mean_squared │
|
179
|
+
# # │ --- ┆ --- ┆ --- │
|
180
|
+
# # │ str ┆ i64 ┆ f64 │
|
181
|
+
# # ╞═════╪═══════╪════════════════╡
|
182
|
+
# # │ a ┆ 2 ┆ 17.0 │
|
183
|
+
# # │ c ┆ 3 ┆ 1.0 │
|
184
|
+
# # │ b ┆ 5 ┆ 10.0 │
|
185
|
+
# # └─────┴───────┴────────────────┘
|
186
|
+
#
|
187
|
+
# @example Use keyword arguments to easily name your expression inputs.
|
188
|
+
# df.group_by("a").agg(
|
189
|
+
# b_sum: Polars.sum("b"),
|
190
|
+
# c_mean_squared: (Polars.col("c") ** 2).mean
|
191
|
+
# )
|
192
|
+
# # =>
|
193
|
+
# # shape: (3, 3)
|
194
|
+
# # ┌─────┬───────┬────────────────┐
|
195
|
+
# # │ a ┆ b_sum ┆ c_mean_squared │
|
196
|
+
# # │ --- ┆ --- ┆ --- │
|
197
|
+
# # │ str ┆ i64 ┆ f64 │
|
198
|
+
# # ╞═════╪═══════╪════════════════╡
|
199
|
+
# # │ a ┆ 2 ┆ 17.0 │
|
200
|
+
# # │ c ┆ 3 ┆ 1.0 │
|
201
|
+
# # │ b ┆ 5 ┆ 10.0 │
|
202
|
+
# # └─────┴───────┴────────────────┘
|
203
|
+
def agg(*aggs, **named_aggs)
|
204
|
+
@df.lazy
|
205
|
+
.group_by(@by, maintain_order: @maintain_order)
|
206
|
+
.agg(*aggs, **named_aggs)
|
207
|
+
.collect(no_optimization: true)
|
208
|
+
end
|
209
|
+
|
210
|
+
# Get the first `n` rows of each group.
|
211
|
+
#
|
212
|
+
# @param n [Integer]
|
213
|
+
# Number of rows to return.
|
214
|
+
#
|
215
|
+
# @return [DataFrame]
|
216
|
+
#
|
217
|
+
# @example
|
218
|
+
# df = Polars::DataFrame.new(
|
219
|
+
# {
|
220
|
+
# "letters" => ["c", "c", "a", "c", "a", "b"],
|
221
|
+
# "nrs" => [1, 2, 3, 4, 5, 6]
|
222
|
+
# }
|
223
|
+
# )
|
224
|
+
# # =>
|
225
|
+
# # shape: (6, 2)
|
226
|
+
# # ┌─────────┬─────┐
|
227
|
+
# # │ letters ┆ nrs │
|
228
|
+
# # │ --- ┆ --- │
|
229
|
+
# # │ str ┆ i64 │
|
230
|
+
# # ╞═════════╪═════╡
|
231
|
+
# # │ c ┆ 1 │
|
232
|
+
# # │ c ┆ 2 │
|
233
|
+
# # │ a ┆ 3 │
|
234
|
+
# # │ c ┆ 4 │
|
235
|
+
# # │ a ┆ 5 │
|
236
|
+
# # │ b ┆ 6 │
|
237
|
+
# # └─────────┴─────┘
|
238
|
+
#
|
239
|
+
# @example
|
240
|
+
# df.group_by("letters").head(2).sort("letters")
|
241
|
+
# # =>
|
242
|
+
# # shape: (5, 2)
|
243
|
+
# # ┌─────────┬─────┐
|
244
|
+
# # │ letters ┆ nrs │
|
245
|
+
# # │ --- ┆ --- │
|
246
|
+
# # │ str ┆ i64 │
|
247
|
+
# # ╞═════════╪═════╡
|
248
|
+
# # │ a ┆ 3 │
|
249
|
+
# # │ a ┆ 5 │
|
250
|
+
# # │ b ┆ 6 │
|
251
|
+
# # │ c ┆ 1 │
|
252
|
+
# # │ c ┆ 2 │
|
253
|
+
# # └─────────┴─────┘
|
254
|
+
def head(n = 5)
|
255
|
+
@df.lazy
|
256
|
+
.group_by(@by, maintain_order: @maintain_order)
|
257
|
+
.head(n)
|
258
|
+
.collect(no_optimization: true)
|
259
|
+
end
|
260
|
+
|
261
|
+
# Get the last `n` rows of each group.
|
262
|
+
#
|
263
|
+
# @param n [Integer]
|
264
|
+
# Number of rows to return.
|
265
|
+
#
|
266
|
+
# @return [DataFrame]
|
267
|
+
#
|
268
|
+
# @example
|
269
|
+
# df = Polars::DataFrame.new(
|
270
|
+
# {
|
271
|
+
# "letters" => ["c", "c", "a", "c", "a", "b"],
|
272
|
+
# "nrs" => [1, 2, 3, 4, 5, 6]
|
273
|
+
# }
|
274
|
+
# )
|
275
|
+
# # =>
|
276
|
+
# # shape: (6, 2)
|
277
|
+
# # ┌─────────┬─────┐
|
278
|
+
# # │ letters ┆ nrs │
|
279
|
+
# # │ --- ┆ --- │
|
280
|
+
# # │ str ┆ i64 │
|
281
|
+
# # ╞═════════╪═════╡
|
282
|
+
# # │ c ┆ 1 │
|
283
|
+
# # │ c ┆ 2 │
|
284
|
+
# # │ a ┆ 3 │
|
285
|
+
# # │ c ┆ 4 │
|
286
|
+
# # │ a ┆ 5 │
|
287
|
+
# # │ b ┆ 6 │
|
288
|
+
# # └─────────┴─────┘
|
289
|
+
#
|
290
|
+
# @example
|
291
|
+
# df.group_by("letters").tail(2).sort("letters")
|
292
|
+
# # =>
|
293
|
+
# # shape: (5, 2)
|
294
|
+
# # ┌─────────┬─────┐
|
295
|
+
# # │ letters ┆ nrs │
|
296
|
+
# # │ --- ┆ --- │
|
297
|
+
# # │ str ┆ i64 │
|
298
|
+
# # ╞═════════╪═════╡
|
299
|
+
# # │ a ┆ 3 │
|
300
|
+
# # │ a ┆ 5 │
|
301
|
+
# # │ b ┆ 6 │
|
302
|
+
# # │ c ┆ 2 │
|
303
|
+
# # │ c ┆ 4 │
|
304
|
+
# # └─────────┴─────┘
|
305
|
+
def tail(n = 5)
|
306
|
+
@df.lazy
|
307
|
+
.group_by(@by, maintain_order: @maintain_order)
|
308
|
+
.tail(n)
|
309
|
+
.collect(no_optimization: true)
|
310
|
+
end
|
311
|
+
|
312
|
+
# Aggregate the first values in the group.
|
313
|
+
#
|
314
|
+
# @return [DataFrame]
|
315
|
+
#
|
316
|
+
# @example
|
317
|
+
# df = Polars::DataFrame.new(
|
318
|
+
# {
|
319
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
320
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
321
|
+
# "c" => [true, true, true, false, false, true],
|
322
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
323
|
+
# }
|
324
|
+
# )
|
325
|
+
# df.group_by("d", maintain_order: true).first
|
326
|
+
# # =>
|
327
|
+
# # shape: (3, 4)
|
328
|
+
# # ┌────────┬─────┬──────┬───────┐
|
329
|
+
# # │ d ┆ a ┆ b ┆ c │
|
330
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
331
|
+
# # │ str ┆ i64 ┆ f64 ┆ bool │
|
332
|
+
# # ╞════════╪═════╪══════╪═══════╡
|
333
|
+
# # │ Apple ┆ 1 ┆ 0.5 ┆ true │
|
334
|
+
# # │ Orange ┆ 2 ┆ 0.5 ┆ true │
|
335
|
+
# # │ Banana ┆ 4 ┆ 13.0 ┆ false │
|
336
|
+
# # └────────┴─────┴──────┴───────┘
|
337
|
+
def first
|
338
|
+
agg(Polars.all.first)
|
339
|
+
end
|
340
|
+
|
341
|
+
# Aggregate the last values in the group.
|
342
|
+
#
|
343
|
+
# @return [DataFrame]
|
344
|
+
#
|
345
|
+
# @example
|
346
|
+
# df = Polars::DataFrame.new(
|
347
|
+
# {
|
348
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
349
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
350
|
+
# "c" => [true, true, true, false, false, true],
|
351
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
352
|
+
# }
|
353
|
+
# )
|
354
|
+
# df.group_by("d", maintain_order: true).last
|
355
|
+
# # =>
|
356
|
+
# # shape: (3, 4)
|
357
|
+
# # ┌────────┬─────┬──────┬───────┐
|
358
|
+
# # │ d ┆ a ┆ b ┆ c │
|
359
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
360
|
+
# # │ str ┆ i64 ┆ f64 ┆ bool │
|
361
|
+
# # ╞════════╪═════╪══════╪═══════╡
|
362
|
+
# # │ Apple ┆ 3 ┆ 10.0 ┆ false │
|
363
|
+
# # │ Orange ┆ 2 ┆ 0.5 ┆ true │
|
364
|
+
# # │ Banana ┆ 5 ┆ 14.0 ┆ true │
|
365
|
+
# # └────────┴─────┴──────┴───────┘
|
366
|
+
def last
|
367
|
+
agg(Polars.all.last)
|
368
|
+
end
|
369
|
+
|
370
|
+
# Reduce the groups to the sum.
|
371
|
+
#
|
372
|
+
# @return [DataFrame]
|
373
|
+
#
|
374
|
+
# @example
|
375
|
+
# df = Polars::DataFrame.new(
|
376
|
+
# {
|
377
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
378
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
379
|
+
# "c" => [true, true, true, false, false, true],
|
380
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
381
|
+
# }
|
382
|
+
# )
|
383
|
+
# df.group_by("d", maintain_order: true).sum
|
384
|
+
# # =>
|
385
|
+
# # shape: (3, 4)
|
386
|
+
# # ┌────────┬─────┬──────┬─────┐
|
387
|
+
# # │ d ┆ a ┆ b ┆ c │
|
388
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
389
|
+
# # │ str ┆ i64 ┆ f64 ┆ u32 │
|
390
|
+
# # ╞════════╪═════╪══════╪═════╡
|
391
|
+
# # │ Apple ┆ 6 ┆ 14.5 ┆ 2 │
|
392
|
+
# # │ Orange ┆ 2 ┆ 0.5 ┆ 1 │
|
393
|
+
# # │ Banana ┆ 9 ┆ 27.0 ┆ 1 │
|
394
|
+
# # └────────┴─────┴──────┴─────┘
|
395
|
+
def sum
|
396
|
+
agg(Polars.all.sum)
|
397
|
+
end
|
398
|
+
|
399
|
+
# Reduce the groups to the minimal value.
|
400
|
+
#
|
401
|
+
# @return [DataFrame]
|
402
|
+
#
|
403
|
+
# @example
|
404
|
+
# df = Polars::DataFrame.new(
|
405
|
+
# {
|
406
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
407
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
408
|
+
# "c" => [true, true, true, false, false, true],
|
409
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"],
|
410
|
+
# }
|
411
|
+
# )
|
412
|
+
# df.group_by("d", maintain_order: true).min
|
413
|
+
# # =>
|
414
|
+
# # shape: (3, 4)
|
415
|
+
# # ┌────────┬─────┬──────┬───────┐
|
416
|
+
# # │ d ┆ a ┆ b ┆ c │
|
417
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
418
|
+
# # │ str ┆ i64 ┆ f64 ┆ bool │
|
419
|
+
# # ╞════════╪═════╪══════╪═══════╡
|
420
|
+
# # │ Apple ┆ 1 ┆ 0.5 ┆ false │
|
421
|
+
# # │ Orange ┆ 2 ┆ 0.5 ┆ true │
|
422
|
+
# # │ Banana ┆ 4 ┆ 13.0 ┆ false │
|
423
|
+
# # └────────┴─────┴──────┴───────┘
|
424
|
+
def min
|
425
|
+
agg(Polars.all.min)
|
426
|
+
end
|
427
|
+
|
428
|
+
# Reduce the groups to the maximal value.
|
429
|
+
#
|
430
|
+
# @return [DataFrame]
|
431
|
+
#
|
432
|
+
# @example
|
433
|
+
# df = Polars::DataFrame.new(
|
434
|
+
# {
|
435
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
436
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
437
|
+
# "c" => [true, true, true, false, false, true],
|
438
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
439
|
+
# }
|
440
|
+
# )
|
441
|
+
# df.group_by("d", maintain_order: true).max
|
442
|
+
# # =>
|
443
|
+
# # shape: (3, 4)
|
444
|
+
# # ┌────────┬─────┬──────┬──────┐
|
445
|
+
# # │ d ┆ a ┆ b ┆ c │
|
446
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
447
|
+
# # │ str ┆ i64 ┆ f64 ┆ bool │
|
448
|
+
# # ╞════════╪═════╪══════╪══════╡
|
449
|
+
# # │ Apple ┆ 3 ┆ 10.0 ┆ true │
|
450
|
+
# # │ Orange ┆ 2 ┆ 0.5 ┆ true │
|
451
|
+
# # │ Banana ┆ 5 ┆ 14.0 ┆ true │
|
452
|
+
# # └────────┴─────┴──────┴──────┘
|
453
|
+
def max
|
454
|
+
agg(Polars.all.max)
|
455
|
+
end
|
456
|
+
|
457
|
+
# Count the number of values in each group.
|
458
|
+
#
|
459
|
+
# @return [DataFrame]
|
460
|
+
#
|
461
|
+
# @example
|
462
|
+
# df = Polars::DataFrame.new(
|
463
|
+
# {
|
464
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
465
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
466
|
+
# "c" => [true, true, true, false, false, true],
|
467
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
468
|
+
# }
|
469
|
+
# )
|
470
|
+
# df.group_by("d", maintain_order: true).count
|
471
|
+
# # =>
|
472
|
+
# # shape: (3, 2)
|
473
|
+
# # ┌────────┬───────┐
|
474
|
+
# # │ d ┆ count │
|
475
|
+
# # │ --- ┆ --- │
|
476
|
+
# # │ str ┆ u32 │
|
477
|
+
# # ╞════════╪═══════╡
|
478
|
+
# # │ Apple ┆ 3 │
|
479
|
+
# # │ Orange ┆ 1 │
|
480
|
+
# # │ Banana ┆ 2 │
|
481
|
+
# # └────────┴───────┘
|
482
|
+
def count
|
483
|
+
agg(Polars.len.alias("count"))
|
484
|
+
end
|
485
|
+
|
486
|
+
# Reduce the groups to the mean values.
|
487
|
+
#
|
488
|
+
# @return [DataFrame]
|
489
|
+
#
|
490
|
+
# @example
|
491
|
+
# df = Polars::DataFrame.new(
|
492
|
+
# {
|
493
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
494
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
495
|
+
# "c" => [true, true, true, false, false, true],
|
496
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
497
|
+
# }
|
498
|
+
# )
|
499
|
+
# df.group_by("d", maintain_order: true).mean
|
500
|
+
# # =>
|
501
|
+
# # shape: (3, 4)
|
502
|
+
# # ┌────────┬─────┬──────────┬──────────┐
|
503
|
+
# # │ d ┆ a ┆ b ┆ c │
|
504
|
+
# # │ --- ┆ --- ┆ --- ┆ --- │
|
505
|
+
# # │ str ┆ f64 ┆ f64 ┆ f64 │
|
506
|
+
# # ╞════════╪═════╪══════════╪══════════╡
|
507
|
+
# # │ Apple ┆ 2.0 ┆ 4.833333 ┆ 0.666667 │
|
508
|
+
# # │ Orange ┆ 2.0 ┆ 0.5 ┆ 1.0 │
|
509
|
+
# # │ Banana ┆ 4.5 ┆ 13.5 ┆ 0.5 │
|
510
|
+
# # └────────┴─────┴──────────┴──────────┘
|
511
|
+
def mean
|
512
|
+
agg(Polars.all.mean)
|
513
|
+
end
|
514
|
+
|
515
|
+
# Count the unique values per group.
|
516
|
+
#
|
517
|
+
# @return [DataFrame]
|
518
|
+
#
|
519
|
+
# @example
|
520
|
+
# df = Polars::DataFrame.new(
|
521
|
+
# {
|
522
|
+
# "a" => [1, 2, 1, 3, 4, 5],
|
523
|
+
# "b" => [0.5, 0.5, 0.5, 10, 13, 14],
|
524
|
+
# "d" => ["Apple", "Banana", "Apple", "Apple", "Banana", "Banana"]
|
525
|
+
# }
|
526
|
+
# )
|
527
|
+
# df.group_by("d", maintain_order: true).n_unique
|
528
|
+
# # =>
|
529
|
+
# # shape: (2, 3)
|
530
|
+
# # ┌────────┬─────┬─────┐
|
531
|
+
# # │ d ┆ a ┆ b │
|
532
|
+
# # │ --- ┆ --- ┆ --- │
|
533
|
+
# # │ str ┆ u32 ┆ u32 │
|
534
|
+
# # ╞════════╪═════╪═════╡
|
535
|
+
# # │ Apple ┆ 2 ┆ 2 │
|
536
|
+
# # │ Banana ┆ 3 ┆ 3 │
|
537
|
+
# # └────────┴─────┴─────┘
|
538
|
+
def n_unique
|
539
|
+
agg(Polars.all.n_unique)
|
540
|
+
end
|
541
|
+
|
542
|
+
# Compute the quantile per group.
|
543
|
+
#
|
544
|
+
# @param quantile [Float]
|
545
|
+
# Quantile between 0.0 and 1.0.
|
546
|
+
# @param interpolation ["nearest", "higher", "lower", "midpoint", "linear"]
|
547
|
+
# Interpolation method.
|
548
|
+
#
|
549
|
+
# @return [DataFrame]
|
550
|
+
#
|
551
|
+
# @example
|
552
|
+
# df = Polars::DataFrame.new(
|
553
|
+
# {
|
554
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
555
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
556
|
+
# "d" => ["Apple", "Orange", "Apple", "Apple", "Banana", "Banana"]
|
557
|
+
# }
|
558
|
+
# )
|
559
|
+
# df.group_by("d", maintain_order: true).quantile(1)
|
560
|
+
# # =>
|
561
|
+
# # shape: (3, 3)
|
562
|
+
# # ┌────────┬─────┬──────┐
|
563
|
+
# # │ d ┆ a ┆ b │
|
564
|
+
# # │ --- ┆ --- ┆ --- │
|
565
|
+
# # │ str ┆ f64 ┆ f64 │
|
566
|
+
# # ╞════════╪═════╪══════╡
|
567
|
+
# # │ Apple ┆ 3.0 ┆ 10.0 │
|
568
|
+
# # │ Orange ┆ 2.0 ┆ 0.5 │
|
569
|
+
# # │ Banana ┆ 5.0 ┆ 14.0 │
|
570
|
+
# # └────────┴─────┴──────┘
|
571
|
+
def quantile(quantile, interpolation: "nearest")
|
572
|
+
agg(Polars.all.quantile(quantile, interpolation: interpolation))
|
573
|
+
end
|
574
|
+
|
575
|
+
# Return the median per group.
|
576
|
+
#
|
577
|
+
# @return [DataFrame]
|
578
|
+
#
|
579
|
+
# @example
|
580
|
+
# df = Polars::DataFrame.new(
|
581
|
+
# {
|
582
|
+
# "a" => [1, 2, 2, 3, 4, 5],
|
583
|
+
# "b" => [0.5, 0.5, 4, 10, 13, 14],
|
584
|
+
# "d" => ["Apple", "Banana", "Apple", "Apple", "Banana", "Banana"]
|
585
|
+
# }
|
586
|
+
# )
|
587
|
+
# df.group_by("d", maintain_order: true).median
|
588
|
+
# # =>
|
589
|
+
# # shape: (2, 3)
|
590
|
+
# # ┌────────┬─────┬──────┐
|
591
|
+
# # │ d ┆ a ┆ b │
|
592
|
+
# # │ --- ┆ --- ┆ --- │
|
593
|
+
# # │ str ┆ f64 ┆ f64 │
|
594
|
+
# # ╞════════╪═════╪══════╡
|
595
|
+
# # │ Apple ┆ 2.0 ┆ 4.0 │
|
596
|
+
# # │ Banana ┆ 4.0 ┆ 13.0 │
|
597
|
+
# # └────────┴─────┴──────┘
|
598
|
+
def median
|
599
|
+
agg(Polars.all.median)
|
600
|
+
end
|
601
|
+
|
602
|
+
# Plot data.
|
603
|
+
#
|
604
|
+
# @return [Vega::LiteChart]
|
605
|
+
def plot(*args, **options)
|
606
|
+
raise ArgumentError, "Multiple groups not supported" if @by.is_a?(::Array) && @by.size > 1
|
607
|
+
# same message as Ruby
|
608
|
+
raise ArgumentError, "unknown keyword: :group" if options.key?(:group)
|
609
|
+
|
610
|
+
@df.plot(*args, **options, group: @by)
|
611
|
+
end
|
612
|
+
end
|
613
|
+
end
|
@@ -0,0 +1,24 @@
|
|
1
|
+
module Polars
|
2
|
+
module IO
|
3
|
+
# Read into a DataFrame from Apache Avro format.
|
4
|
+
#
|
5
|
+
# @param source [Object]
|
6
|
+
# Path to a file or a file-like object.
|
7
|
+
# @param columns [Object]
|
8
|
+
# Columns to select. Accepts a list of column indices (starting at zero) or a list
|
9
|
+
# of column names.
|
10
|
+
# @param n_rows [Integer]
|
11
|
+
# Stop reading from Apache Avro file after reading ``n_rows``.
|
12
|
+
#
|
13
|
+
# @return [DataFrame]
|
14
|
+
def read_avro(source, columns: nil, n_rows: nil)
|
15
|
+
if Utils.pathlike?(source)
|
16
|
+
source = Utils.normalize_filepath(source)
|
17
|
+
end
|
18
|
+
projection, column_names = Utils.handle_projection_columns(columns)
|
19
|
+
|
20
|
+
rbdf = RbDataFrame.read_avro(source, column_names, projection, n_rows)
|
21
|
+
Utils.wrap_df(rbdf)
|
22
|
+
end
|
23
|
+
end
|
24
|
+
end
|