import json import io import pandas as pd import numpy as np import base64 import matplotlib matplotlib.use('Agg') # Wajib dipasang biar server nggak crash import matplotlib.pyplot as plt from django.template.loader import get_template from xhtml2pdf import pisa from sklearn.cluster import KMeans from django.shortcuts import render, redirect, get_object_or_404 from django.contrib.auth import authenticate, login, logout from django.contrib.auth.forms import AuthenticationForm from django.contrib.auth.decorators import login_required from django.contrib import messages from django.db import transaction from django.http import HttpResponse from django.core.paginator import Paginator from .models import SupplyChainData # ================================================================ # 1. AUTHENTICATION & DASHBOARD # ================================================================ def login_view(request): if request.user.is_authenticated: return redirect('dashboard') if request.method == 'POST': form = AuthenticationForm(request, data=request.POST) if form.is_valid(): login(request, form.get_user()) return redirect('dashboard') messages.error(request, "Username atau password salah.") else: form = AuthenticationForm() return render(request, 'analisis/login.html', {'form': form}) def logout_view(request): logout(request) return redirect('login') @login_required def dashboard_view(request): user_data = list(SupplyChainData.objects.filter(user=request.user)) # --------------------------------------------------------- # 🧠 MACHINE LEARNING K-MEANS CORE ENGINE # --------------------------------------------------------- if len(user_data) >= 3: X = np.array([item.quantity for item in user_data]).reshape(-1, 1) kmeans = KMeans(n_clusters=3, random_state=42, n_init=10) labels = kmeans.fit_predict(X) centroids = kmeans.cluster_centers_.flatten() sorted_idx = np.argsort(centroids) label_mapping = { sorted_idx[0]: 'Dead Stock', sorted_idx[1]: 'Slow Moving', sorted_idx[2]: 'Fast Moving' } for i, item in enumerate(user_data): item.cluster_label = label_mapping[labels[i]] else: for item in user_data: item.cluster_label = 'Dead Stock' # --- PENGURUTAN TABEL REKAP (FAST -> SLOW -> DEAD) --- urutan_klaster = {'Fast Moving': 1, 'Slow Moving': 2, 'Dead Stock': 3} user_data.sort(key=lambda x: urutan_klaster.get(x.cluster_label, 4)) # ---------------------------------------------------------- fast_moving = sorted([i for i in user_data if i.cluster_label == 'Fast Moving'], key=lambda x: x.quantity, reverse=True) slow_moving = sorted([i for i in user_data if i.cluster_label == 'Slow Moving'], key=lambda x: x.quantity, reverse=True) dead_stock = sorted([i for i in user_data if i.cluster_label == 'Dead Stock'], key=lambda x: x.quantity) chart_labels_list = [p.product_name for p in fast_moving[:5]] chart_values_list = [float(p.quantity) for p in fast_moving[:5]] # --- DATA PROMO DUMMY BUAT DASHBOARD --- # Lu bisa ganti teksnya di sini sesuai kebutuhan presentasi skripsi # --- DATA PROMO DUMMY BUAT DASHBOARD --- promo_list = [ {'nama_barang': 'Pristine Botol 600ml (Dus)', 'periode': '01 Des - 31 Des', 'keterangan_diskon': 'Diskon 10%'}, {'nama_barang': 'Aqua Botol 1500ml (Dus)', 'periode': '01 Des - 31 Des', 'keterangan_diskon': 'Diskon 10%'}, {'nama_barang': 'Pristine Botol 400ml (Dus)', 'periode': '01 Des - 31 Des', 'keterangan_diskon': 'Diskon 10%'} ] # -------------------------------------------------------------------------------- context = { 'top_fast': fast_moving[:5], 'top_dead': dead_stock[:5], 'all_items': user_data, 'fast_count': len(fast_moving), 'slow_count': len(slow_moving), 'dead_count': len(dead_stock), 'total_produk': len(user_data), 'total_unit': sum(item.quantity for item in user_data), 'chart_labels': json.dumps(chart_labels_list), 'chart_values': json.dumps(chart_values_list), 'promo_list': promo_list, } return render(request, 'analisis/dashboard.html', context) # ================================================================ # 2. UPLOAD & PROCESSING (ETL ENGINE) # ================================================================ @login_required def upload_view(request): if request.method == 'POST': dataset = request.FILES.get('file_dataset') if not dataset: messages.error(request, "Silakan pilih file terlebih dahulu.") return redirect('upload') try: xls = pd.ExcelFile(dataset) # Fungsi Penyapu Ranjau (Cari Baris Judul) def get_clean_dataframe(sheet_name, keyword='Kode'): df_raw = pd.read_excel(xls, sheet_name) mask = df_raw.apply(lambda row: row.astype(str).str.contains(keyword, case=False, na=False).any(), axis=1) if mask.any(): idx = mask.idxmax() df_clean = df_raw.iloc[idx + 1:].copy() df_clean.columns = df_raw.iloc[idx].astype(str).str.strip() df_clean = df_clean.loc[:, df_clean.columns != 'nan'] return df_clean df_raw.columns = df_raw.columns.astype(str).str.strip() return df_raw df_master = get_clean_dataframe('Master Produk', 'Kode') df_stok = get_clean_dataframe('Laporan Stok Gudang', 'Kode Barang') df_beli = get_clean_dataframe('Data Pembelian', 'Kode Barang') df_jual = get_clean_dataframe('Data Penjualan', 'Kode Barang') if 'Kode Produk' in df_master.columns: df_master = df_master.rename(columns={'Kode Produk': 'Kode Barang'}) for df in [df_master, df_stok, df_beli, df_jual]: df.dropna(subset=['Kode Barang'], inplace=True) df['Kode Barang'] = df['Kode Barang'].astype(str).str.strip().str.upper() saldo_awal = df_stok.groupby('Kode Barang')['Stok Awal'].sum().reset_index() col_beli = 'Qty Masuk' if 'Qty Masuk' in df_beli.columns else 'Qty' col_jual = 'Qty Keluar' if 'Qty Keluar' in df_jual.columns else 'Qty' total_beli = df_beli.groupby('Kode Barang')[col_beli].sum().reset_index().rename(columns={col_beli: 'Total Beli'}) jual_grouped = df_jual.groupby('Kode Barang').agg({ col_jual: 'sum', 'Nominal Transaksi': 'sum' }).reset_index().rename(columns={col_jual: 'Total Jual'}) df_final = df_master[['Kode Barang', 'Nama Barang', 'Satuan']].copy() df_final = df_final.merge(saldo_awal, on='Kode Barang', how='left') df_final = df_final.merge(total_beli, on='Kode Barang', how='left') df_final = df_final.merge(jual_grouped, on='Kode Barang', how='left') df_final.fillna(0, inplace=True) df_final['Stok Akhir'] = df_final['Stok Awal'] + df_final['Total Beli'] - df_final['Total Jual'] with transaction.atomic(): SupplyChainData.objects.filter(user=request.user).delete() objs = [] for _, row in df_final.iterrows(): stok_awal_val = int(row['Stok Awal']) if pd.notna(row['Stok Awal']) else 0 pembelian_val = int(row['Total Beli']) if pd.notna(row['Total Beli']) else 0 penjualan_val = int(row['Total Jual']) if pd.notna(row['Total Jual']) else 0 stok_akhir_val = int(row['Stok Akhir']) if pd.notna(row['Stok Akhir']) else 0 objs.append(SupplyChainData( user=request.user, input_kode=str(row['Kode Barang'])[:50], product_name=str(row['Nama Barang'])[:255], category_name=str(row.get('Satuan', 'General'))[:100], sales=float(row['Nominal Transaksi']) if pd.notna(row['Nominal Transaksi']) else 0, quantity=penjualan_val, stok_awal=stok_awal_val, pembelian=pembelian_val, penjualan=penjualan_val, stok_akhir=stok_akhir_val, order_status='Completed' )) SupplyChainData.objects.bulk_create(objs) messages.success(request, f"Sukses! {len(objs)} produk berhasil diimpor beserta mutasi stoknya.") return redirect('dashboard') except Exception as e: messages.error(request, f"Gagal memproses file: {str(e)}") return redirect('upload') return render(request, 'analisis/upload.html') # ================================================================ # 3. INVENTORY LIST & DETAIL # ================================================================ @login_required def inventory_list_view(request): query = request.GET.get('q') all_data = list(SupplyChainData.objects.filter(user=request.user).order_by('-id')) if len(all_data) >= 3: X = np.array([item.quantity for item in all_data]).reshape(-1, 1) kmeans = KMeans(n_clusters=3, random_state=42, n_init=10) labels = kmeans.fit_predict(X) centroids = kmeans.cluster_centers_.flatten() sorted_idx = np.argsort(centroids) label_mapping = { sorted_idx[0]: 'Dead Stock', sorted_idx[1]: 'Slow Moving', sorted_idx[2]: 'Fast Moving' } for i, item in enumerate(all_data): item.cluster_label = label_mapping[labels[i]] else: for item in all_data: item.cluster_label = 'Dead Stock' # --- KODE FIX: Ngurutin dari Fast -> Slow -> Dead --- urutan_klaster = {'Fast Moving': 1, 'Slow Moving': 2, 'Dead Stock': 3} all_data.sort(key=lambda x: urutan_klaster.get(x.cluster_label, 4)) # ---------------------------------------------------- if query: filtered_data = [ item for item in all_data if query.lower() in item.product_name.lower() or query.lower() in str(item.input_kode).lower() ] else: filtered_data = all_data paginator = Paginator(filtered_data, 10) page_obj = paginator.get_page(request.GET.get('page')) return render(request, 'analisis/inventory_list.html', {'page_obj': page_obj}) @login_required def inventory_detail_view(request, item_id): all_data = list(SupplyChainData.objects.filter(user=request.user)) item = get_object_or_404(SupplyChainData, id=item_id, user=request.user) if len(all_data) >= 3: X = np.array([i.quantity for i in all_data]).reshape(-1, 1) kmeans = KMeans(n_clusters=3, random_state=42, n_init=10) labels = kmeans.fit_predict(X) centroids = kmeans.cluster_centers_.flatten() sorted_idx = np.argsort(centroids) label_mapping = { sorted_idx[0]: 'Dead Stock', sorted_idx[1]: 'Slow Moving', sorted_idx[2]: 'Fast Moving' } for i, data_obj in enumerate(all_data): if data_obj.id == item.id: item.cluster_label = label_mapping[labels[i]] break else: item.cluster_label = 'Dead Stock' return render(request, 'analisis/inventory_detail.html', {'item': item}) # ================================================================ # 4. PDF EXPORT # ================================================================ @login_required def export_pdf_view(request): user_data = list(SupplyChainData.objects.filter(user=request.user).order_by('-quantity')) if len(user_data) >= 3: X = np.array([item.quantity for item in user_data]).reshape(-1, 1) kmeans = KMeans(n_clusters=3, random_state=42, n_init=10) labels = kmeans.fit_predict(X) centroids = kmeans.cluster_centers_.flatten() sorted_idx = np.argsort(centroids) label_mapping = { sorted_idx[0]: 'Dead Stock', sorted_idx[1]: 'Slow Moving', sorted_idx[2]: 'Fast Moving' } for i, item in enumerate(user_data): item.cluster_label = label_mapping[labels[i]] else: for item in user_data: item.cluster_label = 'Dead Stock' fast_items = sorted([i for i in user_data if i.cluster_label == 'Fast Moving'], key=lambda x: x.quantity, reverse=True) slow_items = sorted([i for i in user_data if i.cluster_label == 'Slow Moving'], key=lambda x: x.quantity, reverse=True) dead_items = sorted([i for i in user_data if i.cluster_label == 'Dead Stock'], key=lambda x: x.quantity) # ------------------------------------------------------------- # 🎨 3A. GRAFIK PIE CHART (PROPORSI STATUS) # ------------------------------------------------------------- sizes = [len(fast_items), len(slow_items), len(dead_items)] pie_labels = ['Fast Moving', 'Slow Moving', 'Dead Stock'] pie_colors = ['#16a34a', '#f59e0b', '#dc2626'] plt.figure(figsize=(5, 3.5)) if sum(sizes) > 0: plt.pie(sizes, labels=pie_labels, colors=pie_colors, autopct='%1.1f%%', startangle=140, textprops={'fontsize': 9, 'weight': 'bold'}) plt.title('Proporsi Status Inventory', fontsize=11, fontweight='bold') buf_pie = io.BytesIO() plt.savefig(buf_pie, format='png', bbox_inches='tight') plt.close() buf_pie.seek(0) pie_base64 = base64.b64encode(buf_pie.getvalue()).decode('utf-8') buf_pie.close() # ------------------------------------------------------------- # 📊 3B. GRAFIK BAR CHART (RESTOCK VS TERJUAL) # ------------------------------------------------------------- target_items = fast_items[:5] + dead_items[:5] bar_labels = [item.product_name[:15] + ('..' if len(item.product_name)>15 else '') for item in target_items] bar_terjual = [item.penjualan for item in target_items] bar_restock = [item.pembelian for item in target_items] x = np.arange(len(bar_labels)) width = 0.35 plt.figure(figsize=(9, 4)) ax = plt.gca() ax.bar(x - width/2, bar_restock, width, label='Restock (Masuk)', color='#3b82f6') ax.bar(x + width/2, bar_terjual, width, label='Terjual (Keluar)', color='#f97316') ax.set_ylabel('Jumlah Unit', fontsize=9) ax.set_title('Komparasi Arus Barang (Top 5 Paling Laris vs Top 5 Paling Numpuk)', fontsize=11, fontweight='bold') ax.set_xticks(x) ax.set_xticklabels(bar_labels, rotation=45, ha="right", fontsize=8) ax.legend(fontsize=8) plt.tight_layout() buf_bar = io.BytesIO() plt.savefig(buf_bar, format='png', bbox_inches='tight') plt.close() buf_bar.seek(0) bar_base64 = base64.b64encode(buf_bar.getvalue()).decode('utf-8') buf_bar.close() context = { 'user': request.user, 'fast_items': fast_items, 'slow_items': slow_items, 'dead_items': dead_items, 'pie_chart': pie_base64, 'bar_chart': bar_base64, } template_path = 'analisis/pdf_report.html' template = get_template(template_path) html = template.render(context) response = HttpResponse(content_type='application/pdf') response['Content-Disposition'] = 'attachment; filename="Laporan_KMeans_Inventory.pdf"' pisa_status = pisa.CreatePDF(html, dest=response) if pisa_status.err: return HttpResponse('We had some errors 
' + html + '
') return response