Squash merge develop into main

dev/ord-dev.py

@@ -0,0 +1,239 @@
+#!/usr/bin/env python
+# coding: utf-8
+
+# In[1]:
+
+
+import pandas as pd
+
+UNIT_CONVERTER = 24.5
+data = pd.read_excel('../media/data/order2024.xlsx', engine="openpyxl")
+
+    # Rename columns
+data = data.rename(columns={
+        "กำหนดส่ง": "due_date",
+        "แผ่นหน้า": "front_sheet",
+        "ลอน C": "c_wave",
+        "แผ่นกลาง": "middle_sheet",
+        "ลอน B": "b_wave",
+        "แผ่นหลัง": "back_sheet",
+        "จน.ชั้น": "level",
+        "กว้างผลิต": "width",
+        "ยาวผลิต": "length",
+        "ทับเส้นซ้าย": "left_edge_cut",
+        "ทับเส้นกลาง": "middle_edge_cut",
+        "ทับเส้นขวา": "right_edge_cut",
+        "เลขที่ใบสั่งขาย": "order_number",
+        "ชนิดส่วนประกอบ": "component_type",
+        "จำนวนสั่งขาย": "quantity",
+        "จำนวนสั่งผลิต": "production_quantity",
+        "ประเภททับเส้น": "edge_type",
+        "สถานะใบสั่ง": "order_status",
+        "% ที่เกิน": "excess_percentage"
+    })
+
+
+
+# In[2]:
+
+
+import uuid
+
+for index, row in data.iterrows():
+    data.at[index, 'width'] = round(row["width"] / UNIT_CONVERTER, 4)
+    data.at[index, 'length'] = round(row["length"] / UNIT_CONVERTER, 4)
+    data.at[index, 'id'] = uuid.uuid4()
+
+
+# In[ ]:
+
+
+
+
+
+# In[3]:
+
+
+start_date = pd.to_datetime('2024-08-1')
+stop_date = pd.to_datetime('2024-08-7')
+
+
+
+
+# In[4]:
+
+
+
+
+
+# In[6]:
+
+
+from icecream import ic
+import numpy as np
+from dataclasses import dataclass
+from typing import Any, Dict
+from numpy import roll
+import pandas as pd
+
+
+
+# In[7]:
+
+
+# from modules.ordplan import ORD
+
+# legacy_data = ORD(filtered_data, deadline_range = 30, size=temp_size).get()
+
+
+
+# In[8]:
+
+from modules.lp import LP
+
+# In[9]:
+
+
+from modules.new_ga import GA
+
+# In[10]:
+
+from modules.hd import HD
+
+# In[11]:
+
+
+from ordplan_project.settings import MAX_TRIM, MIN_TRIM
+from typing import Dict, Any
+
+def is_trim_fit(trim: float):
+    """
+    Check if trim exceed min/max tirm.
+    """
+    return trim <= MAX_TRIM and trim >= MIN_TRIM
+
+
+def handle_switcher(used) -> Dict[str,Any]:
+    if is_trim_fit(used):
+        return
+    switcher = LP({'fitness': used}).run().get()
+    # ic(switcher,used)
+    return switcher
+
+def update_acc_list(acc_list,ga_instance,hd_instance,elapsed_times):
+        temp_acc_str = ""
+        temp_acc_str+='Parameters -> '
+
+        parameters = (
+            "parent:" +ga_instance.parent_selection_type + " | " +
+            "co proba:" +str(ga_instance.crossover_probability) + " | " +
+            "crossover:" +ga_instance.crossover_type + " | " +
+            str(ga_instance.mutation_probability) + " | " +
+            str(ga_instance.mutation_percent_genes) + " | " +
+            "heu type:" +hd_instance.h_type +"\n"
+        )
+
+        temp_acc_str+=parameters
+
+        accuracy = (passed_trim / count) * 100.00 if count != 0 else 0.0
+        temp_acc_str += "Accuracy -> {:.2f}%".format(accuracy)
+
+        total_seconds = sum(elapsed_times)
+
+        minutes, seconds = divmod(total_seconds, 60)
+
+        time_spent_str = f"{int(minutes)}:{int(seconds):02d}"
+
+        temp_acc_str += f"\nTime Spent -> {time_spent_str}"
+
+        temp_acc_str += "\n"
+
+        acc_list.append(temp_acc_str)
+
+
+# In[14]:
+
+
+from tqdm import tqdm
+import itertools
+import time
+
+import enlighten
+
+
+crossover_types = ["two_points", "uniform"]
+mutation_proba = [None,[0.25, 0.05]]
+crossover_proba = [None, 0.65]
+acc_list = []
+
+x = 1
+count = 2
+
+hd_instances = []
+h_types = ["ff", "ffa"]
+for h_type in h_types:
+    hd_instance = HD(orders=data,h_type=h_type, x=x, start_date=start_date, stop_date=stop_date)
+    hd_instances.append(hd_instance)
+
+temp_size = hd_instance.temp_size
+
+manager = enlighten.get_manager()
+types = manager.counter(total=len(list((itertools.product(crossover_types, mutation_proba, crossover_proba, hd_instances)))), desc='Types', unit='it')
+counts = manager.counter(total=count*len(list((itertools.product(crossover_types, mutation_proba, crossover_proba, hd_instances)))), desc='Count', unit='it')
+
+for crossover_prob, crossover, mutation, hd_instance in itertools.product(crossover_proba, crossover_types, mutation_proba,hd_instances):
+    heuristic_data = hd_instance.get()
+    elapsed_times = []
+    passed_trim = 0
+    types.update()
+    for i in range(count):
+        counts.update()
+        start_time = time.time()
+        ga_instance = GA(
+            heuristic_data,
+            size=temp_size,
+            num_generations=50,
+            out_range=5,
+            seed=i,
+            parent_selection_type="tournament",
+            crossover_type=crossover,
+            mutation_probability=mutation,
+            crossover_probability=crossover_prob
+        )
+        ga_instance.run()
+        size = ga_instance.PAPER_SIZE
+        trim = abs(ga_instance.fitness_values)
+        used = ga_instance.PAPER_SIZE+ga_instance.fitness_values
+        switcher = handle_switcher(used)
+
+        if switcher is not None:
+            size = switcher["new_roll"]
+            trim = switcher["new_trim"]
+
+        # print(size)
+        # print(used)
+        # print(trim)
+        if trim < 3:
+            passed_trim+=1
+        end_time = time.time()
+        elapsed_time = end_time - start_time
+        elapsed_times.append(elapsed_time)
+
+
+    update_acc_list(acc_list,ga_instance,hd_instance,elapsed_times)
+
+
+
+
+
+# In[15]:
+
+
+for i in acc_list:
+    print(i)
+
+
+# In[ ]:
+
+
+