File size: 10,773 Bytes
ef71549
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import gradio as gr
import pandas as pd
from typing import List, Dict, Union, Optional

class SmartSelectColumns(gr.SelectColumns):
    """
    Enhanced SelectColumns component that supports substring matching and column mapping.
    Inherits from gr.SelectColumns but adds additional filtering capabilities.
    """
    def __init__(
        self,
        *args,
        column_filters: Optional[Dict[str, List[str]]] = None,
        column_mapping: Optional[Dict[str, str]] = None,
        **kwargs
    ):
        """
        Initialize the SmartSelectColumns component.
        
        Args:
            column_filters: Dict mapping filter names to lists of substrings to match
            column_mapping: Dict mapping display names to actual column names
            *args, **kwargs: Arguments passed to parent SelectColumns
        """
        super().__init__(*args, **kwargs)
        self.column_filters = column_filters or {}
        self.column_mapping = column_mapping or {}
        
    def preprocess(self, x: List[str]) -> List[str]:
        """Transform selected display names back to actual column names."""
        if self.column_mapping:
            reverse_mapping = {v: k for k, v in self.column_mapping.items()}
            return [reverse_mapping.get(col, col) for col in x]
        return x
    
    def get_filtered_columns(self, df: pd.DataFrame) -> Dict[str, List[str]]:
        """
        Get columns filtered by substring matches.
        
        Args:
            df: Input DataFrame
            
        Returns:
            Dict mapping filter names to lists of matching columns
        """
        filtered_cols = {}
        
        for filter_name, substrings in self.column_filters.items():
            matching_cols = []
            for col in df.columns:
                if any(substr.lower() in col.lower() for substr in substrings):
                    matching_cols.append(col)
            filtered_cols[filter_name] = matching_cols
            
        return filtered_cols

    def update(
        self, 
        value: Union[pd.DataFrame, Dict[str, List[str]]], 
        interactive: Optional[bool] = None
    ) -> Dict:
        """
        Update the component with new values.
        
        Args:
            value: Either a DataFrame or dict of predefined column groups
            interactive: Whether the component should be interactive
        
        Returns:
            Dict containing the update configuration
        """
        if isinstance(value, pd.DataFrame):
            # Get filtered column groups
            filtered_cols = self.get_filtered_columns(value)
            
            # Create display names for columns if mapping exists
            choices = list(value.columns)
            if self.column_mapping:
                choices = [self.column_mapping.get(col, col) for col in choices]
                
            return {
                "choices": choices,
                "filtered_cols": filtered_cols,
                "interactive": interactive if interactive is not None else self.interactive
            }
        return super().update(value, interactive)

# Example usage
if __name__ == "__main__":
    df = pd.DataFrame({
        "ioi_score_1": [1, 2, 3],
        "ioi_score_2": [4, 5, 6],
        "other_metric": [7, 8, 9],
        "performance_1": [10, 11, 12]
    })
    
    # Define filters and mappings
    column_filters = {
        "IOI Metrics": ["ioi"],
        "Performance Metrics": ["performance"]
    }
    
    column_mapping = {
        "ioi_score_1": "IOI Score (Type 1)",
        "ioi_score_2": "IOI Score (Type 2)",
        "other_metric": "Other Metric",
        "performance_1": "Performance Metric 1"
    }
    
    # Create interface
    with gr.Blocks() as demo:
        select_cols = SmartSelectColumns(
            column_filters=column_filters,
            column_mapping=column_mapping,
            multiselect=True
        )
        
        # Update component with DataFrame
        select_cols.update(df)
        
    demo.launch()
































import gradio as gr
import pandas as pd
from typing import List, Dict, Union, Optional, Any
from dataclasses import fields

class SmartSelectColumns(gr.SelectColumns):
    """
    Enhanced SelectColumns component for Gradio Leaderboard with smart filtering and mapping capabilities.
    """
    def __init__(
        self,
        column_filters: Optional[Dict[str, List[str]]] = None,
        column_mapping: Optional[Dict[str, str]] = None,
        initial_selected: Optional[List[str]] = None,
        *args,
        **kwargs
    ):
        """
        Initialize SmartSelectColumns with enhanced functionality.
        
        Args:
            column_filters: Dict mapping filter names to lists of substrings to match
            column_mapping: Dict mapping actual column names to display names
            initial_selected: List of column names to be initially selected
            *args, **kwargs: Additional arguments passed to parent SelectColumns
        """
        super().__init__(*args, **kwargs)
        self.column_filters = column_filters or {}
        self.column_mapping = column_mapping or {}
        self.reverse_mapping = {v: k for k, v in self.column_mapping.items()} if column_mapping else {}
        self.initial_selected = initial_selected or []
        
    def preprocess(self, x: List[str]) -> List[str]:
        """
        Transform selected display names back to actual column names.
        
        Args:
            x: List of selected display names
            
        Returns:
            List of actual column names
        """
        return [self.reverse_mapping.get(col, col) for col in x]
    
    def postprocess(self, y: List[str]) -> List[str]:
        """
        Transform actual column names to display names.
        
        Args:
            y: List of actual column names
            
        Returns:
            List of display names
        """
        return [self.column_mapping.get(col, col) for col in y]
    
    def get_filtered_columns(self, df: pd.DataFrame) -> Dict[str, List[str]]:
        """
        Get columns filtered by substring matches.
        
        Args:
            df: Input DataFrame
            
        Returns:
            Dict mapping filter names to lists of matching display names
        """
        filtered_cols = {}
        
        for filter_name, substrings in self.column_filters.items():
            matching_cols = []
            for col in df.columns:
                if any(substr.lower() in col.lower() for substr in substrings):
                    display_name = self.column_mapping.get(col, col)
                    matching_cols.append(display_name)
            filtered_cols[filter_name] = matching_cols
            
        return filtered_cols

    def update(
        self, 
        value: Union[pd.DataFrame, Dict[str, List[str]], Any], 
        interactive: Optional[bool] = None
    ) -> Dict:
        """
        Update component with new values, supporting DataFrame fields.
        
        Args:
            value: DataFrame, dict of columns, or fields object
            interactive: Whether component should be interactive
            
        Returns:
            Dict containing update configuration
        """
        if isinstance(value, pd.DataFrame):
            filtered_cols = self.get_filtered_columns(value)
            choices = [self.column_mapping.get(col, col) for col in value.columns]
            
            # Set initial selection if provided
            value = self.initial_selected if self.initial_selected else choices
            
            return {
                "choices": choices,
                "value": value,
                "filtered_cols": filtered_cols,
                "interactive": interactive if interactive is not None else self.interactive
            }
        
        # Handle fields object (e.g., from dataclass)
        if hasattr(value, '__dataclass_fields__'):
            field_names = [field.name for field in fields(value)]
            choices = [self.column_mapping.get(name, name) for name in field_names]
            return {
                "choices": choices,
                "value": self.initial_selected if self.initial_selected else choices,
                "interactive": interactive if interactive is not None else self.interactive
            }
            
        return super().update(value, interactive)

def initialize_leaderboard(df: pd.DataFrame, column_class: Any, 
                         filters: Dict[str, List[str]], 
                         mappings: Dict[str, str],
                         initial_columns: Optional[List[str]] = None) -> gr.Leaderboard:
    """
    Initialize a Gradio Leaderboard with SmartSelectColumns.
    
    Args:
        df: Input DataFrame
        column_class: Class containing column definitions (e.g., AutoEvalColumn_mib_subgraph)
        filters: Column filters for substring matching
        mappings: Column name mappings (actual -> display)
        initial_columns: List of columns to show initially
        
    Returns:
        Configured Leaderboard instance
    """
    # Create renamed DataFrame with display names
    renamed_df = df.rename(columns=mappings)
    
    # Initialize SmartSelectColumns
    smart_columns = SmartSelectColumns(
        column_filters=filters,
        column_mapping=mappings,
        initial_selected=initial_columns,
        multiselect=True
    )
    
    return gr.Leaderboard(
        value=renamed_df,
        datatype=[c.type for c in fields(column_class)],
        select_columns=smart_columns,
        search_columns=["Method"],
        hide_columns=[],
        interactive=False
    )

# Example usage
if __name__ == "__main__":
    # Sample data
    df = pd.DataFrame({
        "ioi_score_1": [1, 2, 3],
        "ioi_score_2": [4, 5, 6],
        "other_metric": [7, 8, 9],
        "performance_1": [10, 11, 12],
        "Method": ["A", "B", "C"]
    })
    
    # Define filters and mappings
    filters = {
        "IOI Metrics": ["ioi"],
        "Performance Metrics": ["performance"]
    }
    
    mappings = {
        "ioi_score_1": "IOI Score (Type 1)",
        "ioi_score_2": "IOI Score (Type 2)",
        "other_metric": "Other Metric",
        "performance_1": "Performance Metric 1"
    }
    
    # Create demo interface
    with gr.Blocks() as demo:
        # Initialize leaderboard with smart columns
        leaderboard = initialize_leaderboard(
            df=df,
            column_class=None,  # Replace with your actual column class
            filters=filters,
            mappings=mappings,
            initial_columns=["Method", "IOI Score (Type 1)"]
        )
        
    demo.launch()