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
use crate::TEMP_SUFFIX;
use crate::chart::Chart;
use crate::core::data::{ColumnVector, Dataset};
use crate::encode::y::StackMode;
use crate::error::ChartonError;
use crate::mark::Mark;
use ahash::AHashMap;
impl<T: Mark> Chart<T> {
pub(crate) fn transform_bar_data(mut self) -> Result<Self, ChartonError> {
// --- STEP 1: Context Extraction ---
let y_enc = self.encoding.y.as_mut().unwrap();
let agg_op = y_enc.aggregate;
let x_enc = self.encoding.x.as_ref().unwrap();
let color_enc_opt = self.encoding.color.as_ref();
let mut x_field = x_enc.field.clone();
let y_field = y_enc.field.clone();
// Check if we are in Pie mode (empty X field).
// If so, we force Stacked mode to create a circular stack.
let is_pie = x_field.is_empty();
if is_pie {
y_enc.stack = StackMode::Stacked;
x_field = format!("{}_virtual_root__", TEMP_SUFFIX);
}
let color_field = color_enc_opt.map(|ce| &ce.field);
// A color field triggers grouping ONLY if it's different from the X axis field.
let has_grouping_color = if let Some(cf) = color_field {
cf != &x_field
} else {
false
};
// --- STEP 2: Aggregate Data into a Lookup Map ---
// We group raw rows by (X-Category, Color-Category) to prepare for aggregation.
let mut group_map: AHashMap<(String, Option<String>), Vec<usize>> = AHashMap::new();
let row_count = self.data.height();
for i in 0..row_count {
let x_val = if is_pie {
"all".to_string()
} else {
self.data.get_str_or(&x_field, i, "null")
};
let c_val = if has_grouping_color {
color_field.map(|cf| self.data.get_str_or(cf, i, "null"))
} else {
None
};
group_map.entry((x_val, c_val)).or_default().push(i);
}
let y_col = self.data.column(&y_field)?;
let mut lookup: AHashMap<(String, Option<String>), f64> = group_map
.into_iter()
.map(|(key, indices)| (key, agg_op.aggregate_by_index(y_col, &indices)))
.collect();
// --- STEP 3: Normalization (100% Stacked / Percentage mode) ---
// If normalization is requested, we divide each value by the sum of its X-group.
// This is essential for "Normalized Rose Charts" where all petals have the same radius.
if y_enc.normalize || y_enc.stack == StackMode::Normalize {
let mut x_sums: AHashMap<String, f64> = AHashMap::new();
for ((x, _), val) in &lookup {
*x_sums.entry(x.clone()).or_insert(0.0) += val;
}
for ((x, _), val) in lookup.iter_mut() {
let sum = x_sums.get(x).cloned().unwrap_or(0.0);
*val = if sum != 0.0 { *val / sum } else { 0.0 };
}
}
// --- STEP 4: Cartesian Product & Gap Filling ---
// We ensure every X-category has the same set of Color-categories (filling missing gaps with 0.0).
// This prevents "shifting" in stacked charts (including Nightingale Rose) when data is sparse.
let x_uniques = if is_pie {
vec!["all".to_string()]
} else {
self.data.column(&x_field)?.unique_values()
};
let c_uniques = if has_grouping_color {
self.data.column(color_field.unwrap())?.unique_values()
} else {
vec![]
};
let mut final_x = Vec::new();
let mut final_y = Vec::new();
let mut final_color = Vec::new();
for x in &x_uniques {
if has_grouping_color {
for c in &c_uniques {
// Using .get() instead of .remove() to keep the lookup intact if needed for debug.
// Missing combinations are filled with 0.0 to maintain stack alignment.
let val = lookup
.get(&(x.clone(), Some(c.clone())))
.cloned()
.unwrap_or(0.0);
final_x.push(x.clone());
final_color.push(c.clone());
final_y.push(val);
}
} else {
let val = lookup.get(&(x.clone(), None)).cloned().unwrap_or(0.0);
final_x.push(x.clone());
final_y.push(val);
}
}
// --- STEP 5: Rebuild Dataset ---
let mut new_ds = Dataset::new();
// 1. Resolve column names and metadata
let x_col_name = if is_pie { "" } else { &x_field };
let total_c = if has_grouping_color {
c_uniques.len()
} else {
1
};
let total_rows = final_x.len();
// 2. Generate layout helper columns (consistent with Boxplot/Errorbar)
// These columns allow the renderer to calculate 'dodge' offsets without re-grouping.
let mut f_groups_count = Vec::with_capacity(total_rows);
let mut f_sub_idx = Vec::with_capacity(total_rows);
// Since final_x/final_color were built using nested loops in STEP 4,
// we replicate that structure here to align helper values.
for _ in &x_uniques {
for j in 0..total_c {
f_groups_count.push(total_c as f64);
f_sub_idx.push(j as f64);
}
}
// 3. Assemble the New Dataset
// Primary Axis (X)
new_ds.add_column(
x_col_name,
ColumnVector::String {
data: final_x,
validity: None,
},
)?;
// Measures (Y)
new_ds.add_column(&y_field, ColumnVector::F64 { data: final_y })?;
// Aesthetic Grouping (Color)
if has_grouping_color {
new_ds.add_column(
color_field.unwrap(),
ColumnVector::String {
data: final_color,
validity: None,
},
)?;
}
// Layout Helpers (The "Secret Sauce" for unified rendering)
new_ds.add_column(
format!("{}_groups_count", TEMP_SUFFIX),
ColumnVector::F64 {
data: f_groups_count,
},
)?;
new_ds.add_column(
format!("{}_sub_idx", TEMP_SUFFIX),
ColumnVector::F64 { data: f_sub_idx },
)?;
// --- STEP 6: Finalization ---
self.data = new_ds;
Ok(self)
}
}