Upload 3 files
Browse files- README.md +7 -7
- app.py +263 -202
- requirements.txt +2 -1
README.md
CHANGED
|
@@ -4,16 +4,16 @@ emoji: 🎲
|
|
| 4 |
colorFrom: pink
|
| 5 |
colorTo: blue
|
| 6 |
sdk: gradio
|
| 7 |
-
sdk_version: "4.44.
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
# Ward Ranking Cleaner & Random Assigner (Gradio)
|
| 13 |
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
3
|
| 18 |
-
|
| 19 |
-
|
|
|
|
| 4 |
colorFrom: pink
|
| 5 |
colorTo: blue
|
| 6 |
sdk: gradio
|
| 7 |
+
sdk_version: "4.44.1"
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
# Ward Ranking Cleaner & Random Assigner (Gradio)
|
| 13 |
|
| 14 |
+
- Auto-detect column mapping (Thai/English keywords + fuzzy)
|
| 15 |
+
- Or map by **column numbers** based on the "Available columns" list
|
| 16 |
+
- Clean to keep only `NAME`, `ID`, and selected ward ranking columns (parse ranks → ints)
|
| 17 |
+
- Assign students by rank round (1→2→3…) with random tie-breaking, respecting **capacity**
|
| 18 |
+
- Pre-check: `#students <= total capacity` (shortage allowed, **not exceed**)
|
| 19 |
+
|
app.py
CHANGED
|
@@ -5,17 +5,24 @@ import numpy as np
|
|
| 5 |
import re
|
| 6 |
from io import BytesIO
|
| 7 |
from typing import List, Dict, Tuple, Optional
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
|
| 9 |
-
APP_TITLE = "Ward Ranking Cleaner & Random Assigner (
|
| 10 |
DESCRIPTION = """
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
-
|
| 18 |
-
|
|
|
|
|
|
|
| 19 |
"""
|
| 20 |
|
| 21 |
WARD_CHOICES = [
|
|
@@ -29,6 +36,20 @@ WARD_CHOICES = [
|
|
| 29 |
("Obstetrics", "สูติศาสตร์"),
|
| 30 |
]
|
| 31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
def read_table(file) -> Tuple[Optional[pd.DataFrame], str]:
|
| 33 |
if file is None:
|
| 34 |
return None, "กรุณาอัปโหลดไฟล์ก่อน (.csv หรือ .xlsx)"
|
|
@@ -46,40 +67,16 @@ def read_table(file) -> Tuple[Optional[pd.DataFrame], str]:
|
|
| 46 |
return None, "รองรับเฉพาะ .csv หรือ .xlsx เท่านั้น"
|
| 47 |
except Exception as e:
|
| 48 |
return None, f"อ่านไฟล์ไม่สำเร็จ: {e}"
|
| 49 |
-
# ปรับชื่อคอลัมน์ (trim)
|
| 50 |
df.columns = [str(c).strip() for c in df.columns]
|
| 51 |
return df, ""
|
| 52 |
|
| 53 |
-
def
|
| 54 |
-
""
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
คืนชื่อคอลัมน์จริงถ้าพบ (ตัวแรกที่พบ), ไม่งั้นคืน None
|
| 59 |
-
"""
|
| 60 |
-
cols = list(df.columns)
|
| 61 |
-
if not flexible:
|
| 62 |
-
return key if key in cols else None
|
| 63 |
-
# โหมดยืดหยุ่น
|
| 64 |
-
# ถ้า key เป็นสตริงธรรมดา ให้ค้นหาแบบ "มี key เป็นส่วนหนึ่งของชื่อคอลัมน์" (case-insensitive)
|
| 65 |
-
try:
|
| 66 |
-
pattern = re.compile(key, flags=re.IGNORECASE)
|
| 67 |
-
for c in cols:
|
| 68 |
-
if re.search(pattern, c):
|
| 69 |
-
return c
|
| 70 |
-
except re.error:
|
| 71 |
-
# ถ้า regex ไม่ valid ให้ fallback เป็น contains (case-insensitive)
|
| 72 |
-
low = key.lower()
|
| 73 |
-
for c in cols:
|
| 74 |
-
if low in c.lower():
|
| 75 |
-
return c
|
| 76 |
-
return None
|
| 77 |
|
| 78 |
def parse_rank(value) -> Optional[int]:
|
| 79 |
-
"""
|
| 80 |
-
รับค่าจากคอลัมน์อันดับ เช่น '1st', 'อันดับ 3', '2', 'third' (จะไม่รองรับคำภาษาอังกฤษเต็ม)
|
| 81 |
-
คืนเป็น int ถ้าพบเลข, ถ้าไม่พบคืน None
|
| 82 |
-
"""
|
| 83 |
if pd.isna(value):
|
| 84 |
return None
|
| 85 |
s = str(value)
|
|
@@ -91,102 +88,127 @@ def parse_rank(value) -> Optional[int]:
|
|
| 91 |
return None
|
| 92 |
return None
|
| 93 |
|
| 94 |
-
def
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
|
| 98 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 99 |
"""
|
| 100 |
-
|
| 101 |
-
|
| 102 |
"""
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 108 |
missing = []
|
| 109 |
-
if name_col
|
| 110 |
-
if id_col
|
| 111 |
raise ValueError(f"หาไม่พบคอลัมน์บังคับ: {', '.join(missing)}")
|
| 112 |
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
for
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
| 130 |
-
for c in
|
| 131 |
-
if c not in
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
cleaned =
|
| 136 |
-
|
| 137 |
-
# แปลงอันดับเป็น int
|
| 138 |
-
ward_cols = [c for c in cleaned.columns if c not in ("NAME", "ID")]
|
| 139 |
-
for c in ward_cols:
|
| 140 |
-
cleaned[c] = cleaned[c].apply(parse_rank).astype("Int64")
|
| 141 |
-
|
| 142 |
-
# จัดเรียงคอลัมน์
|
| 143 |
-
cleaned = cleaned[["NAME", "ID"] + ward_cols]
|
| 144 |
-
|
| 145 |
-
return cleaned, messages
|
| 146 |
|
| 147 |
def random_assign(cleaned: pd.DataFrame,
|
| 148 |
capacities: Dict[str, int],
|
| 149 |
seed: Optional[int] = None) -> Tuple[pd.DataFrame, pd.DataFrame, Dict[str, int]]:
|
| 150 |
-
"""
|
| 151 |
-
สุ่มจัดสรรแบบรอบเลือกอันดับ: เริ่มจากอันดับ 1 → 2 → 3 → ...
|
| 152 |
-
- ในแต่ละอันดับและแต่ละวอร์ด: ถ้าเกิน capacity ที่เหลือ ให้สุ่มเลือก
|
| 153 |
-
- คืนผลลัพธ์: assignments, not_assigned, leftover_capacities
|
| 154 |
-
"""
|
| 155 |
rng = np.random.default_rng(seed)
|
| 156 |
wards = [w for w in cleaned.columns if w not in ("NAME", "ID")]
|
| 157 |
-
# กำหนด capacity ที่ใช้จริง เฉพาะวอร์ดที่อยู่ในตาราง
|
| 158 |
cap = {w: int(capacities.get(w, 0)) for w in wards}
|
| 159 |
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
choice_no = pd.Series(index=cleaned.index, data=pd.NA, dtype="Int64") # อันดับที่ได้
|
| 163 |
|
| 164 |
-
# หาค่า max rank ที่ปรากฏ (เช่น 1..6)
|
| 165 |
max_rank = 0
|
| 166 |
for w in wards:
|
| 167 |
-
|
| 168 |
-
if pd.notna(
|
| 169 |
-
max_rank = max(max_rank, int(
|
| 170 |
|
| 171 |
-
# วนทีละอันดับ
|
| 172 |
for r in range(1, max_rank + 1):
|
| 173 |
-
# ข้ามถ้าทุกวอร์ดเต็มแล้ว
|
| 174 |
if all(c <= 0 for c in cap.values()):
|
| 175 |
break
|
| 176 |
-
# สำหรับแต่ละวอร์ด
|
| 177 |
for w in wards:
|
| 178 |
if cap[w] <= 0:
|
| 179 |
continue
|
| 180 |
-
# ผู้สมัครที่ยังไม่ได้รับการจัดสรร และเลือกวอร์ดนี้ที่อันดับ r
|
| 181 |
mask = (assigned.isna()) & (cleaned[w] == r)
|
| 182 |
candidates = cleaned.index[mask].tolist()
|
| 183 |
-
if
|
| 184 |
continue
|
| 185 |
if len(candidates) <= cap[w]:
|
| 186 |
pick = candidates
|
| 187 |
else:
|
| 188 |
pick = list(rng.choice(candidates, size=cap[w], replace=False))
|
| 189 |
-
# ทำการจ��ดสรร
|
| 190 |
assigned.loc[pick] = w
|
| 191 |
choice_no.loc[pick] = r
|
| 192 |
cap[w] -= len(pick)
|
|
@@ -196,90 +218,116 @@ def random_assign(cleaned: pd.DataFrame,
|
|
| 196 |
result["ChoiceNumber"] = choice_no
|
| 197 |
|
| 198 |
not_assigned = result[result["AssignedWard"].isna()].copy()
|
| 199 |
-
|
| 200 |
-
result_preview = result.copy()
|
| 201 |
-
result_preview = result_preview.fillna("")
|
| 202 |
|
| 203 |
-
|
| 204 |
|
| 205 |
def update_capacity_table(selected_wards: List[str]) -> pd.DataFrame:
|
| 206 |
rows = []
|
| 207 |
for w, th in WARD_CHOICES:
|
| 208 |
if selected_wards and w in selected_wards:
|
| 209 |
rows.append([w, th, 0])
|
| 210 |
-
if not rows:
|
| 211 |
-
return pd.DataFrame(columns=["Ward", "Thai Name", "Capacity"])
|
| 212 |
return pd.DataFrame(rows, columns=["Ward", "Thai Name", "Capacity"])
|
| 213 |
|
| 214 |
-
def
|
| 215 |
-
rows = [["NAME", ""], ["ID", ""]]
|
| 216 |
-
for w, th in WARD_CHOICES:
|
| 217 |
-
if selected_wards and w in selected_wards:
|
| 218 |
-
rows.append([w, ""])
|
| 219 |
-
return pd.DataFrame(rows, columns=["Field", "Your Column Header (exact or regex)"])
|
| 220 |
-
|
| 221 |
-
def on_clean(file, selected_wards, capacity_df, mapping_df, flexible):
|
| 222 |
-
if not selected_wards:
|
| 223 |
-
return gr.update(value="กรุณาเลือกวอร์ดอย่างน้อย 1", visible=True), None, None
|
| 224 |
-
# อ่านไฟล์
|
| 225 |
df, msg = read_table(file)
|
| 226 |
if df is None:
|
| 227 |
-
return gr.update(value=msg, visible=True), None, None
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 228 |
|
| 229 |
-
|
| 230 |
-
|
| 231 |
-
|
| 232 |
-
|
| 233 |
|
| 234 |
-
|
| 235 |
-
|
|
|
|
|
|
|
| 236 |
|
| 237 |
-
|
| 238 |
-
return gr.update(value="กรุณาใส่หัวคอลัมน์ของ NAME และ ID", visible=True), None, None
|
| 239 |
|
| 240 |
-
|
| 241 |
-
|
| 242 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
|
| 244 |
try:
|
| 245 |
-
cleaned
|
| 246 |
except Exception as e:
|
| 247 |
-
return gr.update(value=f"❌ เกิดข้อผิดพลาด: {e}", visible=True), None, None
|
| 248 |
|
| 249 |
-
info = "✓ Cleaning สำเร็จ"
|
| 250 |
-
if messages:
|
| 251 |
-
info += "\n" + "\n".join(messages)
|
| 252 |
-
|
| 253 |
-
# เตรียมไฟล์ดาวน์โหลด
|
| 254 |
buf = BytesIO()
|
| 255 |
cleaned.to_csv(buf, index=False, encoding="utf-8-sig")
|
| 256 |
buf.seek(0)
|
|
|
|
|
|
|
| 257 |
|
| 258 |
-
|
| 259 |
-
|
| 260 |
-
|
| 261 |
-
|
| 262 |
-
|
| 263 |
if cleaned_preview is None:
|
| 264 |
return status, None, None, None, None
|
| 265 |
|
| 266 |
-
#
|
| 267 |
-
# แต่เรามีเฉพาะ preview; จึง clean ซ้ำเพื่อได้ dataframe เต็ม
|
| 268 |
df, _ = read_table(file)
|
| 269 |
-
|
| 270 |
-
|
| 271 |
-
|
| 272 |
-
|
| 273 |
-
|
| 274 |
-
|
| 275 |
-
|
| 276 |
-
|
| 277 |
-
|
| 278 |
-
|
| 279 |
-
return gr.update(value="กรุณากรอก capacity ก่อน", visible=True), None, None, None, None
|
| 280 |
-
|
| 281 |
-
# ทำให้แน่ใจว่ามีคอลัมน์ตามชื่อที่เราคาด
|
| 282 |
cap_df = capacity_df.copy()
|
|
|
|
|
|
|
| 283 |
cap_df.columns = ["Ward", "Thai Name", "Capacity"]
|
| 284 |
cap_df = cap_df[cap_df["Ward"].isin([c for c in cleaned.columns if c not in ("NAME", "ID")])]
|
| 285 |
cap_map = {}
|
|
@@ -289,39 +337,38 @@ def on_assign(file, selected_wards, capacity_df, mapping_df, flexible, seed):
|
|
| 289 |
except Exception:
|
| 290 |
cap_map[str(row["Ward"])] = 0
|
| 291 |
|
| 292 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 293 |
|
| 294 |
-
|
| 295 |
-
out_all = BytesIO()
|
| 296 |
-
assigned.to_csv(out_all, index=False, encoding="utf-8-sig")
|
| 297 |
-
out_all.seek(0)
|
| 298 |
|
|
|
|
|
|
|
| 299 |
out_un = BytesIO()
|
| 300 |
-
not_assigned.to_csv(out_un, index=False, encoding="utf-8-sig")
|
| 301 |
-
out_un.seek(0)
|
| 302 |
-
|
| 303 |
leftover_text = "ความจุคงเหลือ:\n" + "\n".join([f"- {k}: {v}" for k, v in leftover.items()])
|
| 304 |
|
| 305 |
return status, assigned.head(30), ("assigned.csv", out_all), ("not_assigned.csv", out_un), leftover_text
|
| 306 |
|
| 307 |
-
|
| 308 |
with gr.Blocks(title=APP_TITLE) as demo:
|
| 309 |
gr.Markdown(f"# {APP_TITLE}")
|
| 310 |
gr.Markdown(DESCRIPTION)
|
| 311 |
|
| 312 |
with gr.Row():
|
| 313 |
-
file = gr.File(file_count="single", file_types=[".csv", ".xlsx"], label="
|
| 314 |
|
| 315 |
with gr.Accordion("1) เลือกวอร์ดที่ต้องใช้", open=True):
|
| 316 |
selected_wards = gr.CheckboxGroup(
|
| 317 |
choices=[w for w, _ in WARD_CHOICES],
|
| 318 |
label="เลือกวอร์ด (เลือกได้หลายข้อ)",
|
| 319 |
-
value=["Medical", "Surgical"]
|
| 320 |
-
)
|
| 321 |
-
gr.Markdown(
|
| 322 |
-
"คำแปล (อ้างอิง): " +
|
| 323 |
-
", ".join([f"**{w}** = {th}" for w, th in WARD_CHOICES])
|
| 324 |
)
|
|
|
|
| 325 |
|
| 326 |
with gr.Accordion("2) กำหนด Capacity ต่อวอร์ด", open=True):
|
| 327 |
capacity_df = gr.Dataframe(
|
|
@@ -331,46 +378,60 @@ with gr.Blocks(title=APP_TITLE) as demo:
|
|
| 331 |
col_count=3,
|
| 332 |
interactive=True,
|
| 333 |
wrap=True,
|
| 334 |
-
label="
|
| 335 |
)
|
| 336 |
selected_wards.change(fn=update_capacity_table, inputs=selected_wards, outputs=capacity_df)
|
| 337 |
|
| 338 |
-
with gr.Accordion("3)
|
| 339 |
-
gr.Markdown(
|
| 340 |
-
|
| 341 |
-
|
| 342 |
-
|
| 343 |
-
|
| 344 |
-
|
| 345 |
-
|
| 346 |
-
|
| 347 |
-
|
| 348 |
-
|
| 349 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 350 |
|
| 351 |
with gr.Row():
|
| 352 |
-
clean_btn = gr.Button("Clean data (ดูพรีวิว)")
|
| 353 |
-
|
| 354 |
|
| 355 |
-
|
| 356 |
-
preview = gr.Dataframe(label="พรีวิวข้อมูลที่ผ่านการ clean (แสดงหัว 30 แถว)", visible=True)
|
| 357 |
cleaned_file = gr.File(label="ดาวน์โหลดไฟล์ cleaned.csv")
|
| 358 |
-
assigned_preview = gr.Dataframe(label="ตัวอย่างผลการจัดสรร (หัว 30 แถว)", visible=True)
|
| 359 |
-
assigned_file = gr.File(label="ดาวน์โหลดไฟล์ assigned.csv")
|
| 360 |
-
not_assigned_file = gr.File(label="ดาวน์โหลดไฟล์ not_assigned.csv")
|
| 361 |
-
leftover_text = gr.Textbox(label="สรุปความจุคงเหลือ", interactive=False)
|
| 362 |
-
|
| 363 |
-
seed = gr.Textbox(label="Random seed (เว้นว่างเพื่อให้สุ่มใหม่ทุกครั้ง)", value="")
|
| 364 |
|
| 365 |
clean_btn.click(
|
| 366 |
fn=on_clean,
|
| 367 |
-
inputs=[file, selected_wards, capacity_df,
|
| 368 |
-
|
|
|
|
| 369 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 370 |
assign_btn.click(
|
| 371 |
fn=on_assign,
|
| 372 |
-
inputs=[file, selected_wards, capacity_df,
|
| 373 |
-
|
|
|
|
| 374 |
)
|
| 375 |
|
| 376 |
if __name__ == "__main__":
|
|
|
|
| 5 |
import re
|
| 6 |
from io import BytesIO
|
| 7 |
from typing import List, Dict, Tuple, Optional
|
| 8 |
+
try:
|
| 9 |
+
from rapidfuzz import process as rf_process
|
| 10 |
+
HAS_FUZZ = True
|
| 11 |
+
except Exception:
|
| 12 |
+
HAS_FUZZ = False
|
| 13 |
|
| 14 |
+
APP_TITLE = "Ward Ranking Cleaner & Random Assigner (Auto-map + Number Mapping)"
|
| 15 |
DESCRIPTION = """
|
| 16 |
+
**Flow**
|
| 17 |
+
1) อัปโหลดไฟล์ .csv/.xlsx
|
| 18 |
+
2) เลือกวอร์ดที่ใช้ + ใส่ capacity
|
| 19 |
+
3) ตรวจหัวคอลัมน์ที่อ่านได้ (Available columns)
|
| 20 |
+
4) **เลือกวิธี mapping**:
|
| 21 |
+
- Auto-detect (คำไทย/อังกฤษ + fuzzy) → ระบบเติมให้อัตโนมัติ
|
| 22 |
+
- หรือกรอก **หมายเลขคอลัมน์** ตามรายการ Available columns (เลขเริ่ม 1)
|
| 23 |
+
5) Clean → เหลือเฉพาะ NAME, ID, และคอลัมน์วอร์ดที่เลือก (ค่าจัดอันดับถูกแปลงเป็นตัวเลข)
|
| 24 |
+
6) Assign → สุ่มตามลำดับอันดับ โดยเคารพ capacity
|
| 25 |
+
- **จะตรวจว่าจำนวนนักศึกษา <= ผลรวม capacity** (ขาดได้ แต่ห้ามเกิน)
|
| 26 |
"""
|
| 27 |
|
| 28 |
WARD_CHOICES = [
|
|
|
|
| 36 |
("Obstetrics", "สูติศาสตร์"),
|
| 37 |
]
|
| 38 |
|
| 39 |
+
# Keyword dictionary for auto mapping
|
| 40 |
+
AUTO_MAP = {
|
| 41 |
+
"NAME": ["ชื่อ-สกุล", "ชื่อ - สกุล", "fullname", "full name", "name", "student name"],
|
| 42 |
+
"ID": ["รหัสนักศึกษา", "รหัส", "student id", "id", "studentid"],
|
| 43 |
+
"Medical": ["อายุรศาสตร์", "medical"],
|
| 44 |
+
"Medical_1": ["อายุรศาสตร์_1", "medical_1", "med_1"],
|
| 45 |
+
"Medical_2": ["อายุรศาสตร์_2", "medical_2", "med_2"],
|
| 46 |
+
"Surgical": ["ศัลยศาสตร์", "surgical", "surgery"],
|
| 47 |
+
"Pediatric": ["เด็ก", "pediatric", "pediatrics"],
|
| 48 |
+
"Community": ["ชุมชน", "community"],
|
| 49 |
+
"Psychiatric": ["จิตเวช", "psychiatric"],
|
| 50 |
+
"Obstetrics": ["สูติศาสตร์", "obstetrics", "obgyn", "ob/gyn"],
|
| 51 |
+
}
|
| 52 |
+
|
| 53 |
def read_table(file) -> Tuple[Optional[pd.DataFrame], str]:
|
| 54 |
if file is None:
|
| 55 |
return None, "กรุณาอัปโหลดไฟล์ก่อน (.csv หรือ .xlsx)"
|
|
|
|
| 67 |
return None, "รองรับเฉพาะ .csv หรือ .xlsx เท่านั้น"
|
| 68 |
except Exception as e:
|
| 69 |
return None, f"อ่านไฟล์ไม่สำเร็จ: {e}"
|
|
|
|
| 70 |
df.columns = [str(c).strip() for c in df.columns]
|
| 71 |
return df, ""
|
| 72 |
|
| 73 |
+
def available_columns_text(df: pd.DataFrame) -> str:
|
| 74 |
+
lines = ["Available columns:"]
|
| 75 |
+
for i, c in enumerate(df.columns, start=1):
|
| 76 |
+
lines.append(f"{i}. {c}")
|
| 77 |
+
return "\n".join(lines)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
|
| 79 |
def parse_rank(value) -> Optional[int]:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
if pd.isna(value):
|
| 81 |
return None
|
| 82 |
s = str(value)
|
|
|
|
| 88 |
return None
|
| 89 |
return None
|
| 90 |
|
| 91 |
+
def auto_map_columns(df: pd.DataFrame, selected_wards: List[str]) -> Dict[str, int]:
|
| 92 |
+
"""Return mapping as index (1-based) for NAME, ID, and selected ward columns.
|
| 93 |
+
Use keyword dictionary and fuzzy fallback (if available)."""
|
| 94 |
+
cols = list(df.columns)
|
| 95 |
+
col_lower = [c.lower() for c in cols]
|
| 96 |
+
result: Dict[str, int] = {}
|
| 97 |
+
|
| 98 |
+
def find_by_keywords(keywords: List[str]) -> Optional[int]:
|
| 99 |
+
for kw in keywords:
|
| 100 |
+
kw_low = kw.lower()
|
| 101 |
+
# contains search
|
| 102 |
+
for idx, c_low in enumerate(col_lower):
|
| 103 |
+
if kw_low in c_low:
|
| 104 |
+
return idx + 1 # 1-based
|
| 105 |
+
# fuzzy fallback
|
| 106 |
+
if HAS_FUZZ:
|
| 107 |
+
best_idx = None
|
| 108 |
+
best_score = -1
|
| 109 |
+
for idx, c in enumerate(cols):
|
| 110 |
+
for kw in keywords:
|
| 111 |
+
match = rf_process.extractOne(kw, [c], score_cutoff=85)
|
| 112 |
+
if match:
|
| 113 |
+
_, score, _ = match
|
| 114 |
+
if score > best_score:
|
| 115 |
+
best_score = score
|
| 116 |
+
best_idx = idx + 1
|
| 117 |
+
if best_idx is not None:
|
| 118 |
+
return best_idx
|
| 119 |
+
return None
|
| 120 |
+
|
| 121 |
+
# NAME / ID
|
| 122 |
+
n_idx = find_by_keywords(AUTO_MAP["NAME"])
|
| 123 |
+
if n_idx: result["NAME"] = n_idx
|
| 124 |
+
i_idx = find_by_keywords(AUTO_MAP["ID"])
|
| 125 |
+
if i_idx: result["ID"] = i_idx
|
| 126 |
+
|
| 127 |
+
# wards
|
| 128 |
+
for w in selected_wards:
|
| 129 |
+
kws = AUTO_MAP.get(w, [w])
|
| 130 |
+
w_idx = find_by_keywords(kws)
|
| 131 |
+
if w_idx:
|
| 132 |
+
result[w] = w_idx
|
| 133 |
+
|
| 134 |
+
return result
|
| 135 |
+
|
| 136 |
+
def build_cleaned_from_indices(df: pd.DataFrame,
|
| 137 |
+
mapping_indices: Dict[str, int]) -> pd.DataFrame:
|
| 138 |
"""
|
| 139 |
+
mapping_indices: {Field -> 1-based column index in df}
|
| 140 |
+
Keep only NAME, ID, and ward columns. Convert ward values to Int (ranks).
|
| 141 |
"""
|
| 142 |
+
# Resolve names
|
| 143 |
+
def idx_to_name(k: str) -> str:
|
| 144 |
+
idx = mapping_indices.get(k, None)
|
| 145 |
+
if idx is None: return ""
|
| 146 |
+
if not (1 <= idx <= len(df.columns)): return ""
|
| 147 |
+
return df.columns[idx - 1]
|
| 148 |
+
|
| 149 |
+
name_col = idx_to_name("NAME")
|
| 150 |
+
id_col = idx_to_name("ID")
|
| 151 |
+
if not name_col or not id_col:
|
| 152 |
missing = []
|
| 153 |
+
if not name_col: missing.append("NAME")
|
| 154 |
+
if not id_col: missing.append("ID")
|
| 155 |
raise ValueError(f"หาไม่พบคอลัมน์บังคับ: {', '.join(missing)}")
|
| 156 |
|
| 157 |
+
# collect ward columns
|
| 158 |
+
ward_cols_src = []
|
| 159 |
+
ward_cols_dst = []
|
| 160 |
+
for w, _th in WARD_CHOICES:
|
| 161 |
+
if w in mapping_indices:
|
| 162 |
+
c = idx_to_name(w)
|
| 163 |
+
if c:
|
| 164 |
+
ward_cols_src.append(c)
|
| 165 |
+
ward_cols_dst.append(w)
|
| 166 |
+
|
| 167 |
+
keep_cols = [name_col, id_col] + ward_cols_src
|
| 168 |
+
cleaned = df[keep_cols].copy()
|
| 169 |
+
rename_map = {name_col: "NAME", id_col: "ID"}
|
| 170 |
+
rename_map.update({src: dst for src, dst in zip(ward_cols_src, ward_cols_dst)})
|
| 171 |
+
cleaned = cleaned.rename(columns=rename_map)
|
| 172 |
+
|
| 173 |
+
# parse ranks
|
| 174 |
+
for c in cleaned.columns:
|
| 175 |
+
if c not in ("NAME", "ID"):
|
| 176 |
+
cleaned[c] = cleaned[c].apply(parse_rank).astype("Int64")
|
| 177 |
+
# order
|
| 178 |
+
ordered = ["NAME", "ID"] + [c for c in cleaned.columns if c not in ("NAME", "ID")]
|
| 179 |
+
cleaned = cleaned[ordered]
|
| 180 |
+
return cleaned
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 181 |
|
| 182 |
def random_assign(cleaned: pd.DataFrame,
|
| 183 |
capacities: Dict[str, int],
|
| 184 |
seed: Optional[int] = None) -> Tuple[pd.DataFrame, pd.DataFrame, Dict[str, int]]:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
rng = np.random.default_rng(seed)
|
| 186 |
wards = [w for w in cleaned.columns if w not in ("NAME", "ID")]
|
|
|
|
| 187 |
cap = {w: int(capacities.get(w, 0)) for w in wards}
|
| 188 |
|
| 189 |
+
assigned = pd.Series(index=cleaned.index, data=pd.NA, dtype="object")
|
| 190 |
+
choice_no = pd.Series(index=cleaned.index, data=pd.NA, dtype="Int64")
|
|
|
|
| 191 |
|
|
|
|
| 192 |
max_rank = 0
|
| 193 |
for w in wards:
|
| 194 |
+
m = cleaned[w].max(skipna=True)
|
| 195 |
+
if pd.notna(m):
|
| 196 |
+
max_rank = max(max_rank, int(m))
|
| 197 |
|
|
|
|
| 198 |
for r in range(1, max_rank + 1):
|
|
|
|
| 199 |
if all(c <= 0 for c in cap.values()):
|
| 200 |
break
|
|
|
|
| 201 |
for w in wards:
|
| 202 |
if cap[w] <= 0:
|
| 203 |
continue
|
|
|
|
| 204 |
mask = (assigned.isna()) & (cleaned[w] == r)
|
| 205 |
candidates = cleaned.index[mask].tolist()
|
| 206 |
+
if not candidates:
|
| 207 |
continue
|
| 208 |
if len(candidates) <= cap[w]:
|
| 209 |
pick = candidates
|
| 210 |
else:
|
| 211 |
pick = list(rng.choice(candidates, size=cap[w], replace=False))
|
|
|
|
| 212 |
assigned.loc[pick] = w
|
| 213 |
choice_no.loc[pick] = r
|
| 214 |
cap[w] -= len(pick)
|
|
|
|
| 218 |
result["ChoiceNumber"] = choice_no
|
| 219 |
|
| 220 |
not_assigned = result[result["AssignedWard"].isna()].copy()
|
| 221 |
+
return result.fillna(""), not_assigned.fillna(""), cap
|
|
|
|
|
|
|
| 222 |
|
| 223 |
+
# ===== Gradio callbacks =====
|
| 224 |
|
| 225 |
def update_capacity_table(selected_wards: List[str]) -> pd.DataFrame:
|
| 226 |
rows = []
|
| 227 |
for w, th in WARD_CHOICES:
|
| 228 |
if selected_wards and w in selected_wards:
|
| 229 |
rows.append([w, th, 0])
|
|
|
|
|
|
|
| 230 |
return pd.DataFrame(rows, columns=["Ward", "Thai Name", "Capacity"])
|
| 231 |
|
| 232 |
+
def on_upload(file, selected_wards):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 233 |
df, msg = read_table(file)
|
| 234 |
if df is None:
|
| 235 |
+
return gr.update(value=msg, visible=True), "", None, None, None
|
| 236 |
+
# Show available columns
|
| 237 |
+
avail = available_columns_text(df)
|
| 238 |
+
# Auto-detect mapping (indices)
|
| 239 |
+
auto_idx = auto_map_columns(df, selected_wards or [])
|
| 240 |
+
# Prepare number inputs defaults
|
| 241 |
+
def idx_or_blank(key):
|
| 242 |
+
return int(auto_idx[key]) if key in auto_idx else None
|
| 243 |
+
name_num = idx_or_blank("NAME")
|
| 244 |
+
id_num = idx_or_blank("ID")
|
| 245 |
+
ward_nums = {w: idx_or_blank(w) for w, _ in WARD_CHOICES}
|
| 246 |
+
return gr.update(value="✓ อ่านไฟล์สำเร็จ", visible=True), avail, name_num, id_num, ward_nums
|
| 247 |
+
|
| 248 |
+
def collect_mapping_numbers(name_num, id_num, ward_nums, selected_wards, n_cols):
|
| 249 |
+
"""Validate numeric mapping and build mapping dict {Field: index}"""
|
| 250 |
+
errors = []
|
| 251 |
+
mapping = {}
|
| 252 |
+
def valid(num, label):
|
| 253 |
+
if num is None:
|
| 254 |
+
errors.append(f"- กรุณาใส่หมายเลขของ {label}")
|
| 255 |
+
return None
|
| 256 |
+
try:
|
| 257 |
+
num = int(num)
|
| 258 |
+
except Exception:
|
| 259 |
+
errors.append(f"- {label} ต้องเป็นตัวเลข")
|
| 260 |
+
return None
|
| 261 |
+
if not (1 <= num <= n_cols):
|
| 262 |
+
errors.append(f"- {label} ต้องอยู่ระหว่าง 1–{n_cols}")
|
| 263 |
+
return None
|
| 264 |
+
return num
|
| 265 |
|
| 266 |
+
nn = valid(name_num, "NAME")
|
| 267 |
+
ii = valid(id_num, "ID")
|
| 268 |
+
if nn: mapping["NAME"] = nn
|
| 269 |
+
if ii: mapping["ID"] = ii
|
| 270 |
|
| 271 |
+
for w in selected_wards:
|
| 272 |
+
wn = valid(ward_nums.get(w, None), f"{w}")
|
| 273 |
+
if wn:
|
| 274 |
+
mapping[w] = wn
|
| 275 |
|
| 276 |
+
return errors, mapping
|
|
|
|
| 277 |
|
| 278 |
+
def on_clean(file, selected_wards, capacity_df, name_num, id_num,
|
| 279 |
+
med_num, med1_num, med2_num, surg_num, ped_num, comm_num, psy_num, obs_num):
|
| 280 |
+
if not selected_wards:
|
| 281 |
+
return gr.update(value="กรุณาเลือกวอร์ดอย่างน้อย 1", visible=True), None, None, None
|
| 282 |
+
|
| 283 |
+
df, msg = read_table(file)
|
| 284 |
+
if df is None:
|
| 285 |
+
return gr.update(value=msg, visible=True), None, None, None
|
| 286 |
+
|
| 287 |
+
n_cols = len(df.columns)
|
| 288 |
+
ward_nums = {
|
| 289 |
+
"Medical": med_num, "Medical_1": med1_num, "Medical_2": med2_num,
|
| 290 |
+
"Surgical": surg_num, "Pediatric": ped_num, "Community": comm_num,
|
| 291 |
+
"Psychiatric": psy_num, "Obstetrics": obs_num
|
| 292 |
+
}
|
| 293 |
+
errors, mapping_idx = collect_mapping_numbers(name_num, id_num, ward_nums, selected_wards, n_cols)
|
| 294 |
+
if errors:
|
| 295 |
+
return gr.update(value="❌ Mapping ไม่ครบ/ไม่ถูกต้อง:\n" + "\n".join(errors), visible=True), None, None, None
|
| 296 |
|
| 297 |
try:
|
| 298 |
+
cleaned = build_cleaned_from_indices(df, mapping_idx)
|
| 299 |
except Exception as e:
|
| 300 |
+
return gr.update(value=f"❌ เกิดข้อผิดพลาด: {e}", visible=True), None, None, None
|
| 301 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 302 |
buf = BytesIO()
|
| 303 |
cleaned.to_csv(buf, index=False, encoding="utf-8-sig")
|
| 304 |
buf.seek(0)
|
| 305 |
+
info = "✓ Cleaning สำเร็จ"
|
| 306 |
+
return gr.update(value=info, visible=True), cleaned.head(30), ("cleaned.csv", buf), len(cleaned)
|
| 307 |
|
| 308 |
+
def on_assign(file, selected_wards, capacity_df, name_num, id_num,
|
| 309 |
+
med_num, med1_num, med2_num, surg_num, ped_num, comm_num, psy_num, obs_num, seed):
|
| 310 |
+
# Clean first to get the cleaned df and student count
|
| 311 |
+
status, cleaned_preview, cleaned_file, n_students = on_clean(file, selected_wards, capacity_df, name_num, id_num,
|
| 312 |
+
med_num, med1_num, med2_num, surg_num, ped_num, comm_num, psy_num, obs_num)
|
| 313 |
if cleaned_preview is None:
|
| 314 |
return status, None, None, None, None
|
| 315 |
|
| 316 |
+
# Recreate full cleaned df (not just head) for assignment
|
|
|
|
| 317 |
df, _ = read_table(file)
|
| 318 |
+
n_cols = len(df.columns)
|
| 319 |
+
ward_nums = {
|
| 320 |
+
"Medical": med_num, "Medical_1": med1_num, "Medical_2": med2_num,
|
| 321 |
+
"Surgical": surg_num, "Pediatric": ped_num, "Community": comm_num,
|
| 322 |
+
"Psychiatric": psy_num, "Obstetrics": obs_num
|
| 323 |
+
}
|
| 324 |
+
_errors, mapping_idx = collect_mapping_numbers(name_num, id_num, ward_nums, selected_wards, n_cols)
|
| 325 |
+
cleaned = build_cleaned_from_indices(df, mapping_idx)
|
| 326 |
+
|
| 327 |
+
# Build capacity map
|
|
|
|
|
|
|
|
|
|
| 328 |
cap_df = capacity_df.copy()
|
| 329 |
+
if cap_df is None or cap_df.empty:
|
| 330 |
+
return gr.update(value="กรุณากรอก capacity ก่อน", visible=True), None, None, None, None
|
| 331 |
cap_df.columns = ["Ward", "Thai Name", "Capacity"]
|
| 332 |
cap_df = cap_df[cap_df["Ward"].isin([c for c in cleaned.columns if c not in ("NAME", "ID")])]
|
| 333 |
cap_map = {}
|
|
|
|
| 337 |
except Exception:
|
| 338 |
cap_map[str(row["Ward"])] = 0
|
| 339 |
|
| 340 |
+
total_capacity = sum(cap_map.values())
|
| 341 |
+
# Pre-check: students must be <= total capacity (ขาดได้แต่ห้ามเกิน)
|
| 342 |
+
if n_students is None:
|
| 343 |
+
n_students = len(cleaned)
|
| 344 |
+
if n_students > total_capacity:
|
| 345 |
+
msg = f"❌ จำนวนผู้สมัคร {n_students} คน มากกว่า capacity รวม {total_capacity} ที่กำหนด (ขาดได้แต่ห้ามเกิน)"
|
| 346 |
+
return gr.update(value=msg, visible=True), None, None, None, None
|
| 347 |
|
| 348 |
+
assigned, not_assigned, leftover = random_assign(cleaned, cap_map, seed=int(seed) if str(seed).strip().isdigit() else None)
|
|
|
|
|
|
|
|
|
|
| 349 |
|
| 350 |
+
out_all = BytesIO()
|
| 351 |
+
assigned.to_csv(out_all, index=False, encoding="utf-8-sig"); out_all.seek(0)
|
| 352 |
out_un = BytesIO()
|
| 353 |
+
not_assigned.to_csv(out_un, index=False, encoding="utf-8-sig"); out_un.seek(0)
|
|
|
|
|
|
|
| 354 |
leftover_text = "ความจุคงเหลือ:\n" + "\n".join([f"- {k}: {v}" for k, v in leftover.items()])
|
| 355 |
|
| 356 |
return status, assigned.head(30), ("assigned.csv", out_all), ("not_assigned.csv", out_un), leftover_text
|
| 357 |
|
|
|
|
| 358 |
with gr.Blocks(title=APP_TITLE) as demo:
|
| 359 |
gr.Markdown(f"# {APP_TITLE}")
|
| 360 |
gr.Markdown(DESCRIPTION)
|
| 361 |
|
| 362 |
with gr.Row():
|
| 363 |
+
file = gr.File(file_count="single", file_types=[".csv", ".xlsx"], label="อัปโหลดข้อมูล (.csv/.xlsx)")
|
| 364 |
|
| 365 |
with gr.Accordion("1) เลือกวอร์ดที่ต้องใช้", open=True):
|
| 366 |
selected_wards = gr.CheckboxGroup(
|
| 367 |
choices=[w for w, _ in WARD_CHOICES],
|
| 368 |
label="เลือกวอร์ด (เลือกได้หลายข้อ)",
|
| 369 |
+
value=["Medical", "Surgical", "Pediatric", "Community", "Psychiatric", "Obstetrics"]
|
|
|
|
|
|
|
|
|
|
|
|
|
| 370 |
)
|
| 371 |
+
gr.Markdown("คำแปล: " + ", ".join([f"**{w}** = {th}" for w, th in WARD_CHOICES]))
|
| 372 |
|
| 373 |
with gr.Accordion("2) กำหนด Capacity ต่อวอร์ด", open=True):
|
| 374 |
capacity_df = gr.Dataframe(
|
|
|
|
| 378 |
col_count=3,
|
| 379 |
interactive=True,
|
| 380 |
wrap=True,
|
| 381 |
+
label="กรอกเฉพาะแถวของวอร์ดที่เลือก"
|
| 382 |
)
|
| 383 |
selected_wards.change(fn=update_capacity_table, inputs=selected_wards, outputs=capacity_df)
|
| 384 |
|
| 385 |
+
with gr.Accordion("3) ตรวจหัวคอลัมน์ & เลือก mapping (Auto/ตัวเลข)", open=True):
|
| 386 |
+
status = gr.Markdown(visible=False)
|
| 387 |
+
available = gr.Code(label="Available columns (เลขเริ่มที่ 1)", language="markdown", interactive=False)
|
| 388 |
+
auto_btn = gr.Button("อ่านไฟล์ & Auto-detect mapping")
|
| 389 |
+
# numeric mapping inputs
|
| 390 |
+
name_num = gr.Number(label="หมายเลขคอลัมน์สำหรับ NAME", precision=0)
|
| 391 |
+
id_num = gr.Number(label="หมายเลขคอลัมน์สำหรับ ID", precision=0)
|
| 392 |
+
with gr.Row():
|
| 393 |
+
med_num = gr.Number(label="หมายเลขคอลัมน์ Medical", precision=0)
|
| 394 |
+
med1_num = gr.Number(label="หมายเลขคอลัมน์ Medical_1", precision=0)
|
| 395 |
+
med2_num = gr.Number(label="หมายเลขคอลัมน์ Medical_2", precision=0)
|
| 396 |
+
with gr.Row():
|
| 397 |
+
surg_num = gr.Number(label="หมายเลขคอลัมน์ Surgical", precision=0)
|
| 398 |
+
ped_num = gr.Number(label="หมายเลขคอลัมน์ Pediatric", precision=0)
|
| 399 |
+
comm_num = gr.Number(label="หมายเลขคอลัมน์ Community", precision=0)
|
| 400 |
+
with gr.Row():
|
| 401 |
+
psy_num = gr.Number(label="หมายเลขคอลัมน์ Psychiatric", precision=0)
|
| 402 |
+
obs_num = gr.Number(label="หมายเลขคอลัมน์ Obstetrics", precision=0)
|
| 403 |
+
|
| 404 |
+
auto_btn.click(fn=on_upload, inputs=[file, selected_wards],
|
| 405 |
+
outputs=[status, available, name_num, id_num,
|
| 406 |
+
{"Medical": med_num, "Medical_1": med1_num, "Medical_2": med2_num,
|
| 407 |
+
"Surgical": surg_num, "Pediatric": ped_num, "Community": comm_num,
|
| 408 |
+
"Psychiatric": psy_num, "Obstetrics": obs_num}])
|
| 409 |
|
| 410 |
with gr.Row():
|
| 411 |
+
clean_btn = gr.Button("Clean data (ดูพรีวิว)", variant="primary")
|
| 412 |
+
seed = gr.Textbox(label="Random seed (เว้นว่างเพื่อสุ่มใหม่)", value="")
|
| 413 |
|
| 414 |
+
preview = gr.Dataframe(label="พรีวิวข้อมูลที่ผ่านการ clean (หัว 30 แถว)", visible=True)
|
|
|
|
| 415 |
cleaned_file = gr.File(label="ดาวน์โหลดไฟล์ cleaned.csv")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 416 |
|
| 417 |
clean_btn.click(
|
| 418 |
fn=on_clean,
|
| 419 |
+
inputs=[file, selected_wards, capacity_df, name_num, id_num,
|
| 420 |
+
med_num, med1_num, med2_num, surg_num, ped_num, comm_num, psy_num, obs_num],
|
| 421 |
+
outputs=[status, preview, cleaned_file, gr.State()]
|
| 422 |
)
|
| 423 |
+
|
| 424 |
+
assign_btn = gr.Button("Assign (สุ่มตามลำดับอันดับ)")
|
| 425 |
+
assigned_preview = gr.Dataframe(label="ตัวอย่างผลการจัดสรร (หัว 30 แถว)")
|
| 426 |
+
assigned_file = gr.File(label="ดาวน์โหลดไฟล์ assigned.csv")
|
| 427 |
+
not_assigned_file = gr.File(label="ดาวน์โหลดไฟล์ not_assigned.csv")
|
| 428 |
+
leftover_text = gr.Textbox(label="สรุปความจุคงเหลือ", interactive=False)
|
| 429 |
+
|
| 430 |
assign_btn.click(
|
| 431 |
fn=on_assign,
|
| 432 |
+
inputs=[file, selected_wards, capacity_df, name_num, id_num,
|
| 433 |
+
med_num, med1_num, med2_num, surg_num, ped_num, comm_num, psy_num, obs_num, seed],
|
| 434 |
+
outputs=[status, assigned_preview, assigned_file, not_assigned_file, leftover_text]
|
| 435 |
)
|
| 436 |
|
| 437 |
if __name__ == "__main__":
|
requirements.txt
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
-
gradio==4.44.
|
| 2 |
pandas==2.2.2
|
| 3 |
openpyxl==3.1.5
|
| 4 |
numpy==2.0.2
|
|
|
|
|
|
| 1 |
+
gradio==4.44.1
|
| 2 |
pandas==2.2.2
|
| 3 |
openpyxl==3.1.5
|
| 4 |
numpy==2.0.2
|
| 5 |
+
rapidfuzz==3.9.7
|