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
//! Window operations for DataFrame
//!
//! This module provides pandas-like window operations for DataFrames including rolling windows,
//! expanding windows, and exponentially weighted moving operations.
use crate::core::error::{Error, Result};
use crate::dataframe::base::DataFrame;
use crate::series::{Series, WindowExt, WindowOps};
/// Extension trait to add window operations to DataFrame
pub trait DataFrameWindowExt {
/// Apply a rolling window operation to a column.
///
/// Equivalent to [`DataFrameWindowExt::rolling_with_options`] with
/// `min_periods = None` (defaults to `window_size`), `center = false`,
/// and `ddof = 1`.
fn rolling(
&self,
window_size: usize,
column_name: &str,
operation: &str,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
/// Apply a rolling window operation to a column with explicit
/// `min_periods`, `center`, and `ddof` (used by `"std"`/`"var"`) --
/// the string-dispatch equivalent of the options the fluent
/// [`crate::dataframe::enhanced_window::DataFrameRolling`] builder
/// already exposes.
fn rolling_with_options(
&self,
window_size: usize,
column_name: &str,
operation: &str,
min_periods: Option<usize>,
center: bool,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
/// Apply an expanding window operation to a column.
///
/// Equivalent to [`DataFrameWindowExt::expanding_with_options`] with
/// `ddof = 1`.
fn expanding(
&self,
min_periods: usize,
column_name: &str,
operation: &str,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
/// Apply an expanding window operation to a column with an explicit
/// `ddof` (used by `"std"`/`"var"`).
fn expanding_with_options(
&self,
min_periods: usize,
column_name: &str,
operation: &str,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
/// Apply an exponentially weighted moving operation to a column.
///
/// Equivalent to [`DataFrameWindowExt::ewm_with_options`] with
/// `ddof = 1`.
fn ewm(
&self,
column_name: &str,
operation: &str,
span: Option<usize>,
alpha: Option<f64>,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
/// Apply an exponentially weighted moving operation to a column with an
/// explicit `ddof` (used by `"std"`/`"var"`).
fn ewm_with_options(
&self,
column_name: &str,
operation: &str,
span: Option<usize>,
alpha: Option<f64>,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame>;
}
impl DataFrameWindowExt for DataFrame {
fn rolling(
&self,
window_size: usize,
column_name: &str,
operation: &str,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
self.rolling_with_options(
window_size,
column_name,
operation,
None,
false,
1,
new_column_name,
)
}
fn rolling_with_options(
&self,
window_size: usize,
column_name: &str,
operation: &str,
min_periods: Option<usize>,
center: bool,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
// Get the column as a Series<f64>
let column = self.get_column_as_f64_legacy(column_name)?;
// Apply the rolling operation
let rolling = column
.rolling(window_size)?
.min_periods(min_periods.unwrap_or(window_size))
.center(center);
let result_series = match operation.to_lowercase().as_str() {
"mean" => rolling.mean()?,
"sum" => rolling.sum()?,
"std" => rolling.std(ddof)?,
"var" => rolling.var(ddof)?,
"min" => rolling.min()?,
"max" => rolling.max()?,
"median" => rolling.median()?,
_ => {
return Err(Error::InvalidValue(format!(
"Unsupported rolling operation: {}",
operation
)))
}
};
// Create a new DataFrame with the result
let col_names: Vec<&str> = self.column_names().iter().map(|s| s.as_str()).collect();
let mut new_df = self.select_columns(&col_names)?;
let default_name = format!("{}_{}", column_name, operation);
let result_column_name = new_column_name.unwrap_or(&default_name);
new_df.add_column(
result_column_name.to_string(),
result_series.to_string_series()?,
)?;
Ok(new_df)
}
fn expanding(
&self,
min_periods: usize,
column_name: &str,
operation: &str,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
self.expanding_with_options(min_periods, column_name, operation, 1, new_column_name)
}
fn expanding_with_options(
&self,
min_periods: usize,
column_name: &str,
operation: &str,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
// Get the column as a Series<f64>
let column = self.get_column_as_f64_legacy(column_name)?;
// Apply the expanding operation
let expanding = column.expanding(min_periods)?;
let result_series = match operation.to_lowercase().as_str() {
"mean" => expanding.mean()?,
"sum" => expanding.sum()?,
"std" => expanding.std(ddof)?,
"var" => expanding.var(ddof)?,
"min" => expanding.min()?,
"max" => expanding.max()?,
"median" => expanding.median()?,
_ => {
return Err(Error::InvalidValue(format!(
"Unsupported expanding operation: {}",
operation
)))
}
};
// Create a new DataFrame with the result
let col_names: Vec<&str> = self.column_names().iter().map(|s| s.as_str()).collect();
let mut new_df = self.select_columns(&col_names)?;
let default_name = format!("{}_{}", column_name, operation);
let result_column_name = new_column_name.unwrap_or(&default_name);
new_df.add_column(
result_column_name.to_string(),
result_series.to_string_series()?,
)?;
Ok(new_df)
}
fn ewm(
&self,
column_name: &str,
operation: &str,
span: Option<usize>,
alpha: Option<f64>,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
self.ewm_with_options(column_name, operation, span, alpha, 1, new_column_name)
}
fn ewm_with_options(
&self,
column_name: &str,
operation: &str,
span: Option<usize>,
alpha: Option<f64>,
ddof: usize,
new_column_name: Option<&str>,
) -> Result<DataFrame> {
// Get the column as a Series<f64>
let column = self.get_column_as_f64_legacy(column_name)?;
// Create EWM window
let mut ewm = column.ewm();
if let Some(span_val) = span {
ewm = ewm.span(span_val);
} else if let Some(alpha_val) = alpha {
ewm = ewm.alpha(alpha_val)?;
} else {
return Err(Error::InvalidValue(
"Must specify either span or alpha for EWM".to_string(),
));
}
// Apply the EWM operation
let result_series = match operation.to_lowercase().as_str() {
"mean" => ewm.mean()?,
"std" => ewm.std(ddof)?,
"var" => ewm.var(ddof)?,
_ => {
return Err(Error::InvalidValue(format!(
"Unsupported EWM operation: {}",
operation
)))
}
};
// Create a new DataFrame with the result
let col_names: Vec<&str> = self.column_names().iter().map(|s| s.as_str()).collect();
let mut new_df = self.select_columns(&col_names)?;
let default_name = format!("{}_{}", column_name, operation);
let result_column_name = new_column_name.unwrap_or(&default_name);
new_df.add_column(
result_column_name.to_string(),
result_series.to_string_series()?,
)?;
Ok(new_df)
}
}
impl DataFrame {
/// Helper method to get a column as `Series<f64>`, regardless of its
/// underlying storage type.
///
/// This used to accept *only* `Series<String>` columns, so a column
/// actually stored as `Series<f64>`/`Series<i64>`/etc. (the normal case
/// for numeric data) always failed with `ColumnNotFound` even though it
/// existed -- silently turning every rolling/expanding/EWM call in this
/// module into a no-op `Ok` for such DataFrames. Delegates to
/// [`DataFrame::get_column_numeric_values`], which downcasts the
/// column's actual element type once instead of assuming a single
/// storage type.
fn get_column_as_f64_legacy(&self, column_name: &str) -> Result<Series<f64>> {
let values = self.get_column_numeric_values(column_name)?;
Series::new(values, Some(column_name.to_string()))
}
}