1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
//! Row operations functionality for OptimizedDataFrame
use rayon::prelude::*;
use std::collections::HashMap;
use super::core::OptimizedDataFrame;
use super::data_ops; // Reference to data operations module
use crate::column::{BooleanColumn, Column, ColumnTrait, Float64Column, Int64Column, StringColumn};
use crate::error::{Error, Result};
use crate::index::{DataFrameIndex, IndexTrait};
impl OptimizedDataFrame {
/// Filter rows (as a new DataFrame)
///
/// Extracts only rows where the value in the condition column (boolean type) is true.
///
/// # Arguments
/// * `condition_column` - Name of the boolean column to use as filter condition
///
/// # Returns
/// * `Result<Self>` - A new DataFrame with filtered rows
///
/// # Note
/// This function has the same signature as the one in the data operations module,
/// so the actual implementation is provided as `filter_rows`.
pub fn filter_rows(&self, condition_column: &str) -> Result<Self> {
// Get condition column
let column_idx = self
.column_indices
.get(condition_column)
.ok_or_else(|| Error::ColumnNotFound(condition_column.to_string()))?;
let condition = &self.columns[*column_idx];
// Verify that the condition column is boolean type
if let Column::Boolean(bool_col) = condition {
// Collect indices of rows where the value is true
let mut indices = Vec::new();
for i in 0..bool_col.len() {
if let Ok(Some(true)) = bool_col.get(i) {
indices.push(i);
}
}
// Create a new DataFrame
let mut result = Self::new();
// Filter each column
for (i, name) in self.column_names.iter().enumerate() {
let column = &self.columns[i];
let filtered_column = match column {
Column::Int64(col) => {
let mut filtered_data = Vec::with_capacity(indices.len());
for &idx in &indices {
if let Ok(Some(val)) = col.get(idx) {
filtered_data.push(val);
} else {
filtered_data.push(0); // Default value
}
}
Column::Int64(Int64Column::new(filtered_data))
}
Column::Float64(col) => {
let mut filtered_data = Vec::with_capacity(indices.len());
for &idx in &indices {
if let Ok(Some(val)) = col.get(idx) {
filtered_data.push(val);
} else {
filtered_data.push(0.0); // Default value
}
}
Column::Float64(Float64Column::new(filtered_data))
}
Column::String(col) => {
let mut filtered_data = Vec::with_capacity(indices.len());
for &idx in &indices {
if let Ok(Some(val)) = col.get(idx) {
filtered_data.push(val.to_string());
} else {
filtered_data.push(String::new()); // Default value
}
}
Column::String(StringColumn::new(filtered_data))
}
Column::Boolean(col) => {
let mut filtered_data = Vec::with_capacity(indices.len());
for &idx in &indices {
if let Ok(Some(val)) = col.get(idx) {
filtered_data.push(val);
} else {
filtered_data.push(false); // Default value
}
}
Column::Boolean(BooleanColumn::new(filtered_data))
}
};
result.add_column(name.clone(), filtered_column)?;
}
// Process the index
if let Some(ref idx) = self.index {
if let DataFrameIndex::Simple(ref simple_idx) = idx {
let mut new_index_values = Vec::with_capacity(indices.len());
for &old_idx in &indices {
if old_idx < simple_idx.len() {
let value = simple_idx
.get_value(old_idx)
.map(|s| s.to_string())
.unwrap_or_else(|| old_idx.to_string());
new_index_values.push(value);
} else {
new_index_values.push(old_idx.to_string());
}
}
let new_index = crate::index::Index::new(new_index_values)?;
result.set_index_from_simple_index(new_index)?;
}
}
Ok(result)
} else {
Err(Error::ColumnTypeMismatch {
name: condition_column.to_string(),
expected: crate::column::ColumnType::Boolean,
found: condition.column_type(),
})
}
}
/// Filter by specified row indices
///
/// # Arguments
/// * `indices` - Array of row indices to extract
///
/// # Returns
/// * `Result<Self>` - A new DataFrame with filtered rows
///
/// # Note
/// This function has the same signature as the one in the data operations module,
/// so the actual implementation is provided as `filter_rows_by_indices`.
pub fn filter_rows_by_indices(&self, indices: &[usize]) -> Result<Self> {
// Use parallel processing
use rayon::prelude::*;
let mut result = Self::new();
// Filter each column in parallel
let column_results: Result<Vec<(String, Column)>> = self
.column_names
.par_iter()
.map(|name| {
let i = self.column_indices[name];
let column = &self.columns[i];
let filtered_column = match column {
Column::Int64(col) => {
let filtered_data: Vec<i64> = indices
.iter()
.filter_map(|&idx| {
if idx < col.len() {
if let Ok(Some(val)) = col.get(idx) {
Some(val)
} else {
Some(0) // Default value
}
} else {
None
}
})
.collect();
Column::Int64(Int64Column::new(filtered_data))
}
Column::Float64(col) => {
let filtered_data: Vec<f64> = indices
.iter()
.filter_map(|&idx| {
if idx < col.len() {
if let Ok(Some(val)) = col.get(idx) {
Some(val)
} else {
Some(0.0) // Default value
}
} else {
None
}
})
.collect();
Column::Float64(Float64Column::new(filtered_data))
}
Column::String(col) => {
let filtered_data: Vec<String> = indices
.iter()
.filter_map(|&idx| {
if idx < col.len() {
if let Ok(Some(val)) = col.get(idx) {
Some(val.to_string())
} else {
Some(String::new()) // Default value
}
} else {
None
}
})
.collect();
Column::String(StringColumn::new(filtered_data))
}
Column::Boolean(col) => {
let filtered_data: Vec<bool> = indices
.iter()
.filter_map(|&idx| {
if idx < col.len() {
if let Ok(Some(val)) = col.get(idx) {
Some(val)
} else {
Some(false) // Default value
}
} else {
None
}
})
.collect();
Column::Boolean(BooleanColumn::new(filtered_data))
}
};
Ok((name.clone(), filtered_column))
})
.collect();
// Process results
let columns = column_results?;
for (name, column) in columns {
result.add_column(name, column)?;
}
// Process the index
if let Some(ref idx) = self.index {
if let DataFrameIndex::Simple(ref simple_idx) = idx {
let valid_indices: Vec<usize> = indices
.iter()
.filter(|&&i| i < self.row_count)
.cloned()
.collect();
if !valid_indices.is_empty() {
let mut new_index_values = Vec::with_capacity(valid_indices.len());
for &old_idx in &valid_indices {
if old_idx < simple_idx.len() {
let value = simple_idx
.get_value(old_idx)
.map(|s| s.to_string())
.unwrap_or_else(|| old_idx.to_string());
new_index_values.push(value);
} else {
new_index_values.push(old_idx.to_string());
}
}
let new_index = crate::index::Index::new(new_index_values)?;
result.set_index_from_simple_index(new_index)?;
}
}
}
Ok(result)
}
/// Get the first n rows
///
/// # Arguments
/// * `n` - Number of rows to retrieve
///
/// # Returns
/// * `Result<Self>` - A new DataFrame with the first n rows
///
/// # Note
/// This function has the same signature as the one in the data operations module,
/// so the actual implementation is provided as `head_rows`.
pub fn head_rows(&self, n: usize) -> Result<Self> {
let n = std::cmp::min(n, self.row_count);
let indices: Vec<usize> = (0..n).collect();
self.filter_rows_by_indices(&indices)
}
/// Get the last n rows
///
/// # Arguments
/// * `n` - Number of rows to retrieve
///
/// # Returns
/// * `Result<Self>` - A new DataFrame with the last n rows
///
/// # Note
/// This function has the same signature as the one in the data operations module,
/// so the actual implementation is provided as `tail_rows`.
pub fn tail_rows(&self, n: usize) -> Result<Self> {
let n = std::cmp::min(n, self.row_count);
let start = self.row_count.saturating_sub(n);
let indices: Vec<usize> = (start..self.row_count).collect();
self.filter_rows_by_indices(&indices)
}
/// Sample rows from the DataFrame
///
/// # Arguments
/// * `n` - Number of rows to sample
/// * `replace` - Whether to sample with replacement
/// * `seed` - Random seed value (for reproducibility)
///
/// # Returns
/// * `Result<Self>` - A new DataFrame with sampled rows
///
/// # Note
/// This function has the same signature as the one in the data operations module,
/// so the actual implementation is provided as `sample_rows`.
pub fn sample_rows(&self, n: usize, replace: bool, seed: Option<u64>) -> Result<Self> {
use rand::rngs::StdRng;
use rand::{seq::SliceRandom, Rng, RngExt, SeedableRng};
if self.row_count == 0 {
return Ok(Self::new());
}
let row_indices: Vec<usize> = (0..self.row_count).collect();
// Initialize random number generator
let mut rng = if let Some(seed_val) = seed {
StdRng::seed_from_u64(seed_val)
} else {
// API changed due to dependency updates, so using a method to generate seed
let mut seed_bytes = [0u8; 32];
rand::rng().fill_bytes(&mut seed_bytes);
StdRng::from_seed(seed_bytes)
};
// Sample row indices
let sampled_indices = if replace {
// Sampling with replacement
let mut samples = Vec::with_capacity(n);
for _ in 0..n {
let idx = rng.random_range(0..self.row_count);
samples.push(idx);
}
samples
} else {
// Sampling without replacement
let sample_size = std::cmp::min(n, self.row_count);
let mut indices_copy = row_indices.clone();
indices_copy.shuffle(&mut rng);
indices_copy[0..sample_size].to_vec()
};
self.filter_rows_by_indices(&sampled_indices)
}
}