`instances` Module¶

Overview¶

The instances module contains code for loading and saving problem instances.

Note

The notation and references (equations, sections, examples, etc.) used below refer to Snyder and Shen, Fundamentals of Supply Chain Theory (FoSCT), 2nd edition (2019).

The abbreviation SCMO below refers to the textbook Supply Chain Modeling and Optimization by Snyder, Smilowitz, and Shen, forthcoming.

Stockpyl has a number of built-in instances, most (but not all) of which are taken from FoSCT. These can be loaded using the load_instance() function by providing the instance name. A list of the built-in instances is provided below.

Deterministic Single-Echelon Inventory Problems

Example 3.1 (EOQ)

Name: example_3_1
Description: Example 3.1 (EOQ)
Code to Load Instance:

instance = load_instance('example_3_1')

Code to Build Manually:

instance = {
    'fixed_cost': 8,
    'holding_cost': 0.225,
    'demand_rate': 1300
}

Problem 3.1 (EOQ)

Name: problem_3_1
Description: Problem 3.1 (EOQ)
Code to Load Instance:

instance = load_instance('problem_3_1')

Code to Build Manually:

instance = {
    'fixed_cost': 2250,
    'holding_cost': 275,
    'demand_rate': 182500
}

Example 3.5 (EOQ w/all-units discounts)

Name: example_3_5
Description: Example 3.5 (EOQ w/all-units discounts)
Code to Load Instance:

instance = load_instance('example_3_5')

Code to Build Manually:

instance = {
    'fixed_cost': 8,
    'holding_cost_rate': 0.3,
    'demand_rate': 1300,
    'breakpoints': [0, 400, 800],
    'unit_costs': [0.75, 0.72, 0.68]
}

Problem 3.12 (EOQ w/all-units discounts)

Name: problem_3_12
Description: Problem 3.12 (EOQ w/all-units discounts)
Code to Load Instance:

instance = load_instance('problem_3_12')

Code to Build Manually:

instance = {
    'fixed_cost': 2250,
    'holding_cost_rate': 0.25,
    'demand_rate': 182500,
    'breakpoints': [0, 1200, 2400],
    'unit_costs': [1490, 1220, 1100]
}

Example 3.8 (EOQB)

Name: example_3_8
Description: Example 3.8 (EOQB)
Code to Load Instance:

instance = load_instance('example_3_8')

Code to Build Manually:

instance = {
    'fixed_cost': 8,
    'holding_cost': 0.225,
    'stockout_cost': 5,
    'demand_rate': 1300
}

Problem 3.2(b) (EOQB)

Name: problem_3_2b
Description: Problem 3.2(b) (EOQB)
Code to Load Instance:

instance = load_instance('problem_3_2b')

Code to Build Manually:

instance = {
    'fixed_cost': 40,
    'holding_cost': 40.05,
    'stockout_cost': 60,
    'demand_rate': 2080
}

Problem 3.22 (EPQ)

Name: problem_3_22
Description: Problem 3.22 (EPQ)
Code to Load Instance:

instance = load_instance('problem_3_22')

Code to Build Manually:

instance = {
    'fixed_cost': 4,
    'holding_cost': 0.08,
    'demand_rate': 80,
    'production_rate': 110
}

Example 3.9 (Wagner-Whitin)

Name: example_3_9
Description: Example 3.9 (Wagner-Whitin)
Code to Load Instance:

instance = load_instance('example_3_9')

Code to Build Manually:

instance = {
    "num_periods": 4,
    "holding_cost": 2,
    "fixed_cost": 500,
    "demand": [
        90,
        120,
        80,
        70
    ]
}

Problem 3.27 (Wagner-Whitin)

Name: problem_3_27
Description: Problem 3.27 (Wagner-Whitin)
Code to Load Instance:

instance = load_instance('problem_3_27')

Code to Build Manually:

instance = {
    "num_periods": 4,
    "holding_cost": 0.8,
    "fixed_cost": 120,
    "demand": [
        150,
        100,
        80,
        200
    ]
}

Problem 3.29 (Wagner-Whitin)

Name: problem_3_29
Description: Problem 3.29 (Wagner-Whitin)
Code to Load Instance:

instance = load_instance('problem_3_29')

Code to Build Manually:

instance = {
    "num_periods": 5,
    "holding_cost": 0.1,
    "fixed_cost": 100,
    "demand": [
        730,
        580,
        445,
        650,
        880
    ]
}

SCMO (Wagner-Whitin with nonstationary purchase cost)

Name: scmo_ww_hw_c
Description: SCMO (Wagner-Whitin with nonstationary purchase cost)
Code to Load Instance:

instance = load_instance('scmo_ww_hw_c')

Code to Build Manually:

instance = {
    "num_periods": 5,
    "holding_cost": 0.1,
    "fixed_cost": 100,
    "demand": [
        400,
        500,
        500,
        1100,
        900
    ],
    "purchase_cost": [
        3,
        1,
        4,
        6,
        6
    ]
}

SCMO (JRP example)

Name: scmo_jrp_ex
Description: SCMO (JRP example)
Code to Load Instance:

instance = load_instance('scmo_jrp_ex')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 600,
    "individual_fixed_costs": [
        120,
        840,
        300
    ],
    "holding_costs": [
        160,
        20,
        50
    ],
    "demand_rates": [
        1,
        1,
        1
    ]
}

Numerical JRP example on p. 428 of Silver, Pyke, and Peterson (1998)

Name: spp_jrp
Description: Numerical JRP example on p. 428 of Silver, Pyke, and Peterson (1998)
Code to Load Instance:

instance = load_instance('spp_jrp')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 40,
    "individual_fixed_costs": [
        15,
        15,
        15,
        15
    ],
    "holding_costs": [
        0.24,
        0.24,
        0.24,
        0.24
    ],
    "demand_rates": [
        86000,
        12500,
        1400,
        3000
    ]
}

Numerical JRP example in Silver (1976)

Name: silver_jrp
Description: Numerical JRP example in Silver (1976)
Code to Load Instance:

instance = load_instance('silver_jrp')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 10,
    "individual_fixed_costs": [
        1.87,
        5.27,
        7.94,
        8.19,
        8.87
    ],
    "holding_costs": [
        0.2,
        0.2,
        0.2,
        0.2,
        0.2
    ],
    "demand_rates": [
        1736,
        656,
        558,
        170,
        142
    ]
}

SCMO (JRP problem 1)

Name: scmo_jrp_hw_1
Description: SCMO (JRP problem 1)
Code to Load Instance:

instance = load_instance('scmo_jrp_hw_1')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 20000,
    "individual_fixed_costs": [
        36000,
        46000,
        34000,
        38000
    ],
    "holding_costs": [
        1000,
        900,
        1200,
        1000
    ],
    "demand_rates": [
        1780,
        445,
        920,
        175
    ]
}

SCMO (JRP problem 2)

Name: scmo_jrp_hw_2
Description: SCMO (JRP problem 2)
Code to Load Instance:

instance = load_instance('scmo_jrp_hw_2')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 1500,
    "individual_fixed_costs": [
        4000,
        1000,
        2000
    ],
    "holding_costs": [
        300,
        200,
        200
    ],
    "demand_rates": [
        9100,
        83200,
        20800
    ]
}

SCMO (JRP problem 3)

Name: scmo_jrp_hw_3
Description: SCMO (JRP problem 3)
Code to Load Instance:

instance = load_instance('scmo_jrp_hw_3')

Code to Build Manually:

instance = {
    "shared_fixed_cost": 180,
    "individual_fixed_costs": [
        60,
        100,
        180,
        115,
        135
    ],
    "holding_costs": [
        5.32,
        3.9200000000000004,
        4.760000000000001,
        3.9200000000000004,
        3.3600000000000003
    ],
    "demand_rates": [
        6200,
        1300,
        400,
        4400,
        1800
    ]
}

Stochastic Single-Echelon Inventory Problems

Example 4.1 (newsvendor)

Name: example_4_1
Description: Example 4.1 (newsvendor)
Code to Load Instance:

instance = load_instance('example_4_1')

Code to Build Manually:

instance = {
    "holding_cost": 0.18,
    "stockout_cost": 0.7,
    "demand_mean": 50,
    "demand_sd": 8
}

Example 4.1 (newsvendor) (as SupplyChainNetwork object)

Name: example_4_1_network
Description: Example 4.1 (newsvendor) (as SupplyChainNetwork object)
Code to Load Instance:

instance = load_instance('example_4_1_network')

Code to Build Manually:

from stockpyl.supply_chain_network import single_stage_system
instance = single_stage_system(
    holding_cost=0.18,
    stockout_cost=0.7,
    demand_type='N',
    mean=50,
    standard_deviation=8,
    shipment_lead_time=1,
    policy_type='BS',
    base_stock_level=56.6
)

Example 4.2 (newsvendor explicit)

Name: example_4_2
Description: Example 4.2 (newsvendor explicit)
Code to Load Instance:

instance = load_instance('example_4_2')

Code to Build Manually:

instance = {
    "revenue": 1,
    "purchase_cost": 0.3,
    "salvage_value": 0.12,
    "demand_mean": 50,
    "demand_sd": 8
}

Example 4.3 (newsvendor) (= Example 4.1)

Name: example_4_3
Description: Example 4.3 (newsvendor) (= Example 4.1)
Code to Load Instance:

instance = load_instance('example_4_3')

Code to Build Manually:

instance = {
    "holding_cost": 0.18,
    "stockout_cost": 0.7,
    "demand_mean": 50,
    "demand_sd": 8
}

Problem 4.1 (newsvendor)

Name: problem_4_1
Description: Problem 4.1 (newsvendor)
Code to Load Instance:

instance = load_instance('problem_4_1')

Code to Build Manually:

instance = {
    "holding_cost": 43,
    "stockout_cost": 79,
    "demand_mean": 900,
    "demand_sd": 60
}

Problem 4.3(b) (newsvendor explicit)

Name: problem_4_3b
Description: Problem 4.3(b) (newsvendor explicit)
Code to Load Instance:

instance = load_instance('problem_4_3b')

Code to Build Manually:

instance = {
    "revenue": 7,
    "purchase_cost": 2.5,
    "salvage_value": 1.5,
    "demand_mean": 50,
    "demand_sd": 10
}

Example 4.4 (base-stock optimization)

Name: example_4_4
Description: Example 4.4 (base-stock optimization)
Code to Load Instance:

instance = load_instance('example_4_4')

Code to Build Manually:

instance = {
    "holding_cost": 0.18,
    "stockout_cost": 0.7,
    "demand_mean": 50,
    "demand_sd": 8,
    "lead_time": 4
}

Example 4.7 ((s,S) with Poisson demand)

Name: example_4_7
Description: Example 4.7 ((s,S) with Poisson demand)
Code to Load Instance:

instance = load_instance('example_4_7')

Code to Build Manually:

instance = {
    "holding_cost": 1,
    "stockout_cost": 4,
    "fixed_cost": 5,
    "demand_mean": 6
}

Problem 4.7(b) (newsvendor with discrete demand)

Name: problem_4_7b
Description: Problem 4.7(b) (newsvendor with discrete demand)
Code to Load Instance:

instance = load_instance('problem_4_7b')

Code to Build Manually:

instance = {
    "holding_cost": 500000,
    "stockout_cost": 1000000,
    "demand_pmf": {
        "1": 0.25,
        "2": 0.05,
        "3": 0.1,
        "4": 0.2,
        "5": 0.15,
        "6": 0.1,
        "7": 0.1,
        "8": 0.05
    }
}

Problem 4.8(a) (newsvendor with Poisson demand)

Name: problem_4_8a
Description: Problem 4.8(a) (newsvendor with Poisson demand)
Code to Load Instance:

instance = load_instance('problem_4_8a')

Code to Build Manually:

instance = {
    "holding_cost": 200,
    "stockout_cost": 270,
    "demand_mean": 18
}

Problem 4.8(b) (newsvendor with lognormal demand)

Name: problem_4_8b
Description: Problem 4.8(b) (newsvendor with lognormal demand)
Code to Load Instance:

instance = load_instance('problem_4_8b')

Code to Build Manually:

instance = {
    "holding_cost": 1,
    "stockout_cost": 0.1765,
    "mu": 6,
    "sigma": 0.3
}

Problem 4.31 ((s,S) with Poisson demand)

Name: problem_4_31
Description: Problem 4.31 ((s,S) with Poisson demand)
Code to Load Instance:

instance = load_instance('problem_4_31')

Code to Build Manually:

instance = {
    "holding_cost": 40,
    "stockout_cost": 125,
    "fixed_cost": 150,
    "demand_mean": 4
}

Example 4.8 ((s,S))

Name: example_4_8
Description: Example 4.8 ((s,S))
Code to Load Instance:

instance = load_instance('example_4_8')

Code to Build Manually:

instance = {
    "holding_cost": 0.18,
    "stockout_cost": 0.7,
    "fixed_cost": 2.5,
    "demand_mean": 50,
    "demand_sd": 8
}

Problem 4.32 ((s,S))

Name: problem_4_32
Description: Problem 4.32 ((s,S))
Code to Load Instance:

instance = load_instance('problem_4_32')

Code to Build Manually:

instance = {
    "holding_cost": 2,
    "stockout_cost": 36,
    "fixed_cost": 60,
    "demand_mean": 190,
    "demand_sd": 48
}

Problem 4.29 (finite-horizon)

Name: problem_4_29
Description: Problem 4.29 (finite-horizon)
Code to Load Instance:

instance = load_instance('problem_4_29')

Code to Build Manually:

instance = {
    "num_periods": 10,
    "holding_cost": 1,
    "stockout_cost": 25,
    "terminal_holding_cost": 1,
    "terminal_stockout_cost": 25,
    "purchase_cost": 1,
    "fixed_cost": 0,
    "demand_mean": 18,
    "demand_sd": 3,
    "discount_factor": 0.98,
    "initial_inventory_level": 0
}

Problem 4.30 (finite-horizon)

Name: problem_4_30
Description: Problem 4.30 (finite-horizon)
Code to Load Instance:

instance = load_instance('problem_4_30')

Code to Build Manually:

instance = {
    "num_periods": 10,
    "holding_cost": 1,
    "stockout_cost": 25,
    "terminal_holding_cost": 1,
    "terminal_stockout_cost": 25,
    "purchase_cost": 1,
    "fixed_cost": 40,
    "demand_mean": 18,
    "demand_sd": 3,
    "discount_factor": 0.98,
    "initial_inventory_level": 0
}

Example 5.1 ((r,Q))

Name: example_5_1
Description: Example 5.1 ((r,Q))
Code to Load Instance:

instance = load_instance('example_5_1')

Code to Build Manually:

instance = {
    "holding_cost": 0.225,
    "stockout_cost": 7.5,
    "fixed_cost": 8,
    "demand_mean": 1300,
    "demand_sd": 150,
    "lead_time": 0.08333333333333333
}

Problem 5.1 ((r,Q))

Name: problem_5_1
Description: Problem 5.1 ((r,Q))
Code to Load Instance:

instance = load_instance('problem_5_1')

Code to Build Manually:

instance = {
    "holding_cost": 3.1,
    "stockout_cost": 45,
    "fixed_cost": 50,
    "demand_mean": 800,
    "demand_sd": 40,
    "lead_time": 0.010958904109589041
}

Problem 5.2 ((r,Q) with Poisson demand)

Name: problem_5_2
Description: Problem 5.2 ((r,Q) with Poisson demand)
Code to Load Instance:

instance = load_instance('problem_5_2')

Code to Build Manually:

instance = {
    "holding_cost": 4,
    "stockout_cost": 28,
    "fixed_cost": 4,
    "demand_mean": 12,
    "lead_time": 0.5
}

Problem 5.3 ((r,Q))

Name: problem_5_3
Description: Problem 5.3 ((r,Q))
Code to Load Instance:

instance = load_instance('problem_5_3')

Code to Build Manually:

instance = {
    "holding_cost": 0.21428571428571427,
    "stockout_cost": 40,
    "fixed_cost": 85,
    "demand_mean": 192,
    "demand_sd": 17.4,
    "lead_time": 3
}

Example 5.8 ((r,Q) with Poisson demand))

Name: example_5_8
Description: Example 5.8 ((r,Q) with Poisson demand))
Code to Load Instance:

instance = load_instance('example_5_8')

Code to Build Manually:

instance = {
    "holding_cost": 20,
    "stockout_cost": 150,
    "fixed_cost": 100,
    "demand_mean": 1.5,
    "lead_time": 2
}

Stochastic Multi-Echelon Inventory Problems

Example 6.1 (serial SSM)

Name: example_6_1
Description: Example 6.1 (serial SSM)
Code to Load Instance:

instance = load_instance('example_6_1')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=3,
    node_order_in_system=[3, 2, 1],
    echelon_holding_cost={1: 3, 2: 2, 3: 2},
    local_holding_cost={1: 7, 2: 4, 3: 2},
    shipment_lead_time={1: 1, 2: 1, 3: 2},
    stockout_cost={1: 37.12, 2: 0, 3: 0},
    demand_type='N',
    mean=5,
    standard_deviation=1,
    policy_type='BS',
    base_stock_level={1: 6.49, 2: 5.53, 3: 10.69}
)

Problem 6.1 (serial SSM)

Name: problem_6_1
Description: Problem 6.1 (serial SSM)
Code to Load Instance:

instance = load_instance('problem_6_1')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=2,
    node_order_in_system=[2, 1],
    node_order_in_lists=[1, 2],
    local_holding_cost=[2, 1],
    echelon_holding_cost=[1, 1],
    stockout_cost=[15, 0],
    demand_type='N',
    mean=100,
    standard_deviation=15,
    shipment_lead_time=[1, 1],
    policy_type='BS',
    base_stock_level=[100, 94]
)

Problem 6.2(a) (serial SSM)

Name: problem_6_2a
Description: Problem 6.2(a) (serial SSM)
Code to Load Instance:

instance = load_instance('problem_6_2a')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=5,
    node_order_in_system=[5, 4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4, 5],
    local_holding_cost=[1, 2, 3, 5, 7],
    echelon_holding_cost=[2, 2, 1, 1, 1],
    stockout_cost=[24, 0, 0, 0, 0],
    demand_type='N',
    mean=64,
    standard_deviation=8,
    shipment_lead_time=[0.5, 0.5, 0.5, 0.5, 0.5],
    policy_type='BS',
    base_stock_level=[40.59, 33.87, 35.14, 33.30, 32.93]
)

Problem 6.2(a) (serial SSM with Poisson demand)

Name: problem_6_2b
Description: Problem 6.2(a) (serial SSM with Poisson demand)
Code to Load Instance:

instance = load_instance('problem_6_2b')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=5,
    node_order_in_system=[5, 4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4, 5],
    local_holding_cost=[1, 2, 3, 5, 7],
    echelon_holding_cost=[2, 2, 1, 1, 1],
    stockout_cost=[24, 0, 0, 0, 0],
    demand_type='P',
    mean=64,
    shipment_lead_time=[0.5, 0.5, 0.5, 0.5, 0.5],
    policy_type='BS',
    base_stock_level=[40.59, 33.87, 35.14, 33.30, 32.93]
)

Problem 6.2(a) (serial SSM), adjusted for periodic review

Name: problem_6_2a_adj
Description: Problem 6.2(a) (serial SSM), adjusted for periodic review
Notes: Since L=0.5 in Problem 6.2(a), here we treat each period as having length 0.5 in the original problem.
Code to Load Instance:

instance = load_instance('problem_6_2a_adj')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
import numpy as np
instance = serial_system(
    num_nodes=5,
    node_order_in_system=[5, 4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4, 5],
    local_holding_cost=list(np.array([1, 2, 3, 5, 7]) / 2),
    stockout_cost=list(np.array([24, 0, 0, 0, 0]) / 2),
    demand_type='N',
    mean=64 / 2,
    standard_deviation=8 / np.sqrt(2),
    shipment_lead_time=[1, 1, 1, 1, 1],
    policy_type='BS',
    base_stock_level=[40.59, 33.87, 35.14, 33.30, 32.93]
)

Problem 6.2(b) (serial SSM with Poisson demand), adjusted for periodic review

Name: problem_6_2b_adj
Description: Problem 6.2(b) (serial SSM with Poisson demand), adjusted for periodic review
Notes: Since L=0.5 in Problem 6.2(b), here we treat each period as having length 0.5 in the original problem.
Code to Load Instance:

instance = load_instance('problem_6_2b_adj')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
import numpy as np
instance = serial_system(
    num_nodes=5,
    node_order_in_system=[5, 4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4, 5],
    local_holding_cost=list(np.array([1, 2, 3, 5, 7]) / 2),
    stockout_cost=list(np.array([24, 0, 0, 0, 0]) / 2),
    demand_type='P',
    mean=64 / 2,
    shipment_lead_time=[1, 1, 1, 1, 1],
    policy_type='BS',
    base_stock_level=[40.59, 33.87, 35.14, 33.30, 32.93]
)

Problem 6.16 (serial SSM)

Name: problem_6_16
Description: Problem 6.16 (serial SSM)
Code to Load Instance:

instance = load_instance('problem_6_16')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=2,
    node_order_in_system=[2, 1],
    node_order_in_lists=[1, 2],
    local_holding_cost=[7, 2],
    echelon_holding_cost=[5, 2],
    stockout_cost=[24, 0],
    demand_type='N',
    mean=20,
    standard_deviation=4,
    shipment_lead_time=[8, 3],
    policy_type='BS',
    base_stock_level=[171.1912, 57.7257],
    initial_inventory_level=20,
    initial_orders=20,
    initial_shipments=20
)

Serial SSM system in Shang and Song (2003), Table 1, row 1

Name: shang_song_1
Description: Serial SSM system in Shang and Song (2003), Table 1, row 1
Code to Load Instance:

instance = load_instance('shang_song_1')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=4,
    node_order_in_system=[4, 3, 2, 1],
    echelon_holding_cost=0.25,
    shipment_lead_time=0.25,
    stockout_cost={1: 9},
    demand_type='P',
    mean=16,
    policy_type='BS'
)

Serial SSM system in Shang and Song (2003), Table 1, row 9

Name: shang_song_9
Description: Serial SSM system in Shang and Song (2003), Table 1, row 9
Code to Load Instance:

instance = load_instance('shang_song_9')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=4,
    node_order_in_system=[4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4],
    echelon_holding_cost=[0.25, 2.5, 2.5, 0.25],
    shipment_lead_time=0.25,
    stockout_cost={1: 9},
    demand_type='P',
    mean=16,
    policy_type='BS'
)

Serial SSM system in Shang and Song (2003), Table 1, row 17

Name: shang_song_17
Description: Serial SSM system in Shang and Song (2003), Table 1, row 17
Code to Load Instance:

instance = load_instance('shang_song_17')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=4,
    node_order_in_system=[4, 3, 2, 1],
    echelon_holding_cost=0.25,
    shipment_lead_time=0.25,
    stockout_cost={1: 99},
    demand_type='P',
    mean=16,
    policy_type='BS'
)

Serial SSM system in Shang and Song (2003), Table 1, row 25

Name: shang_song_25
Description: Serial SSM system in Shang and Song (2003), Table 1, row 25
Code to Load Instance:

instance = load_instance('shang_song_25')

Code to Build Manually:

from stockpyl.supply_chain_network import serial_system
instance = serial_system(
    num_nodes=4,
    node_order_in_system=[4, 3, 2, 1],
    node_order_in_lists=[1, 2, 3, 4],
    echelon_holding_cost=[0.25, 2.5, 2.5, 0.25],
    shipment_lead_time=0.25,
    stockout_cost={1: 99},
    demand_type='P',
    mean=16,
    policy_type='BS'
)

3-stage assembly system (2 warehouses, 1 retailer)

Name: assembly_3_stage
Description: 3-stage assembly system (2 warehouses, 1 retailer)
Code to Load Instance:

instance = load_instance('assembly_3_stage')

Code to Build Manually:

from stockpyl.supply_chain_network import mwor_system
instance = mwor_system(
    num_warehouses=2,
    node_order_in_lists=[0, 1, 2],
    local_holding_cost=[2, 1, 1],
    stockout_cost=[20, 0, 0],
    demand_type='N',
    mean=5,
    standard_deviation=1,
    shipment_lead_time=[1, 2, 2],
    policy_type='BS',
    base_stock_level=[7, 13, 11],
    initial_inventory_level=[7, 13, 11]
)
instance.get_node_from_index(0).demand_source.round_to_int = True

Assembly system from Figure 1 in Rosling (1989)

Name: rosling_figure_1
Description: Assembly system from Figure 1 in Rosling (1989)
Notes: Structure and lead times are from Rosling; all other parameters are made up.
Code to Load Instance:

instance = load_instance('rosling_figure_1')

Code to Build Manually:

from stockpyl.supply_chain_network import SupplyChainNetwork
from stockpyl.supply_chain_node import SupplyChainNode
from stockpyl.demand_source import DemandSource
instance = SupplyChainNetwork()
nodes = {i: SupplyChainNode(index=i) for i in range(1, 8)}
# Inventory policies (balanced echelon base-stock).
for n in nodes.values():
    n.inventory_policy.type = 'BEBS'
    n.inventory_policy.base_stock_level = None
# Node 1.
nodes[1].shipment_lead_time = 1
nodes[1].demand_source = DemandSource(type='UD', lo=0, hi=10)
nodes[1].supply_type = None
nodes[1].inventory_policy.base_stock_level = 8
nodes[1].initial_inventory_level = 8
instance.add_node(nodes[1])
# Node 2.
nodes[2].shipment_lead_time = 1
nodes[2].supply_type = None
nodes[2].inventory_policy.base_stock_level = 24
nodes[2].initial_inventory_level = 8
instance.add_predecessor(nodes[1], nodes[2])
# Node 3.
nodes[3].shipment_lead_time = 3
nodes[3].supply_type = None
nodes[3].inventory_policy.base_stock_level = 40
nodes[3].initial_inventory_level = 24
instance.add_predecessor(nodes[1], nodes[3])
# Node 4.
nodes[4].shipment_lead_time = 2
nodes[4].supply_type = None
nodes[4].inventory_policy.base_stock_level = 76
nodes[4].initial_inventory_level = 16
instance.add_predecessor(nodes[3], nodes[4])
# Node 5.
nodes[5].shipment_lead_time = 4
nodes[5].inventory_policy.base_stock_level = 62
nodes[5].initial_inventory_level = 32
nodes[5].supply_type = 'U'
instance.add_predecessor(nodes[2], nodes[5])
# Node 6.
nodes[6].shipment_lead_time = 1
nodes[6].supply_type = 'U'
nodes[6].inventory_policy.base_stock_level = 84
nodes[6].initial_inventory_level = 8
instance.add_predecessor(nodes[4], nodes[6])
# Node 7.
nodes[7].shipment_lead_time = 2
nodes[7].supply_type = 'U'
nodes[7].inventory_policy.base_stock_level = 92
nodes[7].initial_inventory_level = 16
instance.add_predecessor(nodes[4], nodes[7])

Distribution system from Figure 1(a) in Rong, Atan, and Snyder (2017)

Name: rong_atan_snyder_figure_1a
Description: Distribution system from Figure 1(a) in Rong, Atan, and Snyder (2017)
Notes: Uses normal demand instead of Poisson.
Code to Load Instance:

instance = load_instance('rong_atan_snyder_figure_1a')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
from math import sqrt
instance = network_from_edges(
    edges=[(0, 1), (0, 2), (1, 3), (1, 4), (2, 5), (2, 6)],
    demand_type={0: None, 1: None, 2: None, 3: 'N', 4: 'N', 5: 'N', 6: 'N'},
    mean=8,
    standard_deviation=sqrt(8),
    local_holding_cost={0: 1/3, 1: 2/3, 2: 2/3, 3: 1, 4: 1, 5: 1, 6: 1},
    stockout_cost=20,
    shipment_lead_time=1,
    policy_type='BS',
    base_stock_level={i: 0 for i in range(0, 7)}
)

Distribution system from Figure 1(b) in Rong, Atan, and Snyder (2017)

Name: rong_atan_snyder_figure_1b
Description: Distribution system from Figure 1(b) in Rong, Atan, and Snyder (2017)
Notes: Uses normal demand instead of Poisson. Cost and lead times are omitted.
Code to Load Instance:

instance = load_instance('rong_atan_snyder_figure_1b')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
from math import sqrt
demand_type = {i: 'N' if i >= 3 else None for i in range(11)}
instance = network_from_edges(
    edges=[(0, 1), (0, 2), (1, 3), (1, 4), (1, 5), (1, 6), (1, 7), (2, 8), (2, 9), (2, 10)],
    demand_type=demand_type,
    mean=8,
    standard_deviation=sqrt(8),
    policy_type='BS',
    base_stock_level={i: 0 for i in range(0, 11)}
)

Distribution system from Figure 1(c) in Rong, Atan, and Snyder (2017)

Name: rong_atan_snyder_figure_1c
Description: Distribution system from Figure 1(c) in Rong, Atan, and Snyder (2017)
Notes: Uses normal demand instead of Poisson. Cost and lead times are omitted.
Code to Load Instance:

instance = load_instance('rong_atan_snyder_figure_1c')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
from math import sqrt
instance = network_from_edges(
    edges=[(0, 1), (0, 2), (2, 3), (2, 4), (2, 5)],
    demand_type={0: None, 1: 'N', 2: None, 3: 'N', 4: 'N', 5: 'N'},
    mean=8,
    standard_deviation=sqrt(8),
    policy_type='BS',
    base_stock_level={i: 0 for i in range(0, 6)}
)

Example 6.3 (serial GSM)

Name: example_6_3
Description: Example 6.3 (serial GSM)
Code to Load Instance:

instance = load_instance('example_6_3')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
instance = network_from_edges(
    edges=[(3, 2), (2, 1)],
    node_order_in_lists=[1, 2, 3],
    processing_time=[1, 0, 1],
    external_inbound_cst=[None, None, 1],
    local_holding_cost=[7, 4, 2],
    demand_bound_constant=1,
    external_outbound_cst=[1, None, None],
    demand_type=['N', None, None],
    mean=0,
    standard_deviation=[1, 0, 0]
)

Problem 6.7 (serial GSM)

Name: problem_6_7
Description: Problem 6.7 (serial GSM)
Code to Load Instance:

instance = load_instance('problem_6_7')

Code to Build Manually:

from stockpyl.supply_chain_network import SupplyChainNetwork
from stockpyl.supply_chain_node import SupplyChainNode
from stockpyl.demand_source import DemandSource
instance = SupplyChainNetwork()
instance.add_node(SupplyChainNode(3, name='Forming', network=instance, processing_time=1, local_holding_cost=2, external_inbound_cst=1, demand_bound_constant=4))
instance.add_node(SupplyChainNode(2, name='Firing', network=instance, processing_time=1, local_holding_cost=3, demand_bound_constant=4))
instance.add_node(SupplyChainNode(1, name='Glazing', network=instance, processing_time=2, local_holding_cost=4, external_outbound_cst=0, demand_source=DemandSource(type='N', mean=45, standard_deviation=10), demand_bound_constant=4))
instance.add_edges_from_list([(3, 2), (2, 1)])

Problem 6.8 (serial GSM)

Name: problem_6_8
Description: Problem 6.8 (serial GSM)
Code to Load Instance:

instance = load_instance('problem_6_8')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
from scipy import stats
instance = network_from_edges(
    [(n+1, n) for n in range(1, 10)],
    node_order_in_lists=list(range(1, 11)),
    processing_time=[5, 10, 2, 15, 8, 5, 9, 5, 1, 5],
    external_inbound_cst=[None] * 9 + [7],
    local_holding_cost=[5.73, 4.56, 3.04, 2.93, 2.47, 2.37, 1.15, 1.1, 0.98, 0.87],
    demand_bound_constant=stats.norm.ppf(0.98),
    external_outbound_cst=[3] + [None] * 9,
    demand_type=['N'] + [None] * 9,
    mean=0,
    standard_deviation=[15.8] + [None] * 9
)

Problem 6.9 (tree GSM)

Name: problem_6_9
Description: Problem 6.9 (tree GSM)
Code to Load Instance:

instance = load_instance('problem_6_9')

Code to Build Manually:

from stockpyl.supply_chain_network import SupplyChainNetwork
from stockpyl.supply_chain_node import SupplyChainNode
from stockpyl.demand_source import DemandSource
instance = SupplyChainNetwork()
instance.add_node(SupplyChainNode(1, network=instance, processing_time=7, local_holding_cost=220*0.2/365, external_outbound_cst=3, demand_source=DemandSource(type='N', mean=22.0, standard_deviation=4.1), demand_bound_constant=4))
instance.add_node(SupplyChainNode(2, network=instance, processing_time=7, local_holding_cost=140*0.2/365, external_outbound_cst=3, demand_source=DemandSource(type='N', mean=15.3, standard_deviation=6.2), demand_bound_constant=4))
instance.add_node(SupplyChainNode(3, network=instance, processing_time=21, local_holding_cost=90*0.2/365, demand_bound_constant=4))
instance.add_node(SupplyChainNode(4, network=instance, processing_time=3, local_holding_cost=5*0.2/365, demand_bound_constant=4))
instance.add_node(SupplyChainNode(5, network=instance, processing_time=8, local_holding_cost=20*0.2/365, demand_bound_constant=4))
instance.add_node(SupplyChainNode(6, network=instance, processing_time=2, local_holding_cost=7.5*0.2/365, demand_bound_constant=4))
instance.add_edges_from_list([(6, 5), (4, 3), (5, 3), (3, 1), (3, 2)])

Example 6.5 (tree GSM)

Name: example_6_5
Description: Example 6.5 (tree GSM)
Code to Load Instance:

instance = load_instance('example_6_5')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
instance = network_from_edges(
    edges=[(1, 3), (3, 2), (3, 4)],
    node_order_in_lists=[1, 2, 3, 4],
    processing_time=[2, 1, 1, 1],
    external_inbound_cst=[1, None, None, None],
    local_holding_cost=[1, 3, 2, 3],
    demand_bound_constant=[1, 1, 1, 1],
    external_outbound_cst=[None, 0, None, 1],
    demand_type=[None, 'N', None, 'N'],
    mean=[None, 0, None, 0],
    standard_deviation=[None, 1, None, 1]
)

Figure 6.12 (tree GSM)

Name: figure_6_12
Description: Figure 6.12 (tree GSM)
Notes: This instance only consists of the network structure, no costs or other parameters.
Code to Load Instance:

instance = load_instance('figure_6_12')

Code to Build Manually:

from stockpyl.supply_chain_network import network_from_edges
instance = network_from_edges(
    edges=[(1, 2), (1, 3), (3, 5), (4, 5), (5, 6), (5, 7)]
)

Figure 6.14 (tree GSM)

Name: figure_6_14
Description: Figure 6.14 (tree GSM)
Code to Load Instance:

instance = load_instance('figure_6_14')

Code to Build Manually:

from stockpyl.supply_chain_network import SupplyChainNetwork
from stockpyl.supply_chain_node import SupplyChainNode
from stockpyl.demand_source import DemandSource
from scipy import stats
instance = SupplyChainNetwork()
instance.add_node(SupplyChainNode(1, 'Raw_Material', instance, processing_time=2, local_holding_cost=0.01))
instance.add_node(SupplyChainNode(2, 'Process_Wafers', instance, processing_time=3, local_holding_cost=0.03))
instance.add_node(SupplyChainNode(3, 'Package_Test_Wafers', instance, processing_time=2, local_holding_cost=0.04))
instance.add_node(SupplyChainNode(4, 'Imager_Base', instance, processing_time=4, local_holding_cost=0.06))
instance.add_node(SupplyChainNode(5, 'Imager_Assembly', instance, processing_time=2, local_holding_cost=0.12))
instance.add_node(SupplyChainNode(6, 'Ship_to_Final_Assembly', instance, processing_time=3, local_holding_cost=0.13))
instance.add_node(SupplyChainNode(7, 'Camera', instance, processing_time=6, local_holding_cost=0.20))
instance.add_node(SupplyChainNode(8, 'Circuit_Board', instance, processing_time=4, local_holding_cost=0.08))
instance.add_node(SupplyChainNode(9, 'Other_Parts', instance, processing_time=3, local_holding_cost=0.04))
instance.add_node(SupplyChainNode(10, 'Build_Test_Pack', instance, processing_time=2, local_holding_cost=0.50, \
    external_outbound_cst=2, demand_source=DemandSource(type='N', mean=0, standard_deviation=10)))
for n in instance.nodes:
    n.demand_bound_constant = stats.norm.ppf(0.95)
instance.add_edges_from_list([(1, 2), (2, 3), (3, 5), (4, 5), (5, 6), (7, 10), (6, 10), (8, 10), (9, 10)])

Single-Echelon Inventory Problems with Supply Uncertainty

Example 9.1 (EOQD)

Name: example_9_1
Description: Example 9.1 (EOQD)
Code to Load Instance:

instance = load_instance('example_9_1')

Code to Build Manually:

instance = {
    "holding_cost": 0.225,
    "stockout_cost": 5,
    "fixed_cost": 8,
    "demand_rate": 1300,
    "disruption_rate": 1.5,
    "recovery_rate": 14
}

Problem 9.8 (EOQD)

Name: problem_9_8
Description: Problem 9.8 (EOQD)
Code to Load Instance:

instance = load_instance('problem_9_8')

Code to Build Manually:

instance = {
    "holding_cost": 4,
    "stockout_cost": 22,
    "fixed_cost": 35,
    "demand_rate": 30,
    "disruption_rate": 1,
    "recovery_rate": 12
}

Example 9.3 (base-stock with disruptions)

Name: example_9_3
Description: Example 9.3 (base-stock with disruptions)
Code to Load Instance:

instance = load_instance('example_9_3')

Code to Build Manually:

instance = {
    "holding_cost": 0.25,
    "stockout_cost": 3,
    "demand": 2000,
    "disruption_prob": 0.04,
    "recovery_prob": 0.25
}

Example 9.4 (EOQ with additive yield uncertainty)

Name: example_9_4
Description: Example 9.4 (EOQ with additive yield uncertainty)
Code to Load Instance:

instance = load_instance('example_9_4')

Code to Build Manually:

instance = {
    "holding_cost": 0.06,
    "fixed_cost": 18500,
    "demand_rate": 75000,
    "yield_mean": -15000,
    "yield_sd": 9000
}

Example 9.5 (EOQ with multiplicative yield uncertainty)

Name: example_9_5
Description: Example 9.5 (EOQ with multiplicative yield uncertainty)
Code to Load Instance:

instance = load_instance('example_9_5')

Code to Build Manually:

instance = {
    "holding_cost": 0.06,
    "fixed_cost": 18500,
    "demand_rate": 75000,
    "yield_mean": 0.8333333333333334,
    "yield_sd": 0.14085904245475275
}

Problem 9.4(a) (EOQ with additive yield uncertainty)

Name: problem_9_4a
Description: Problem 9.4(a) (EOQ with additive yield uncertainty)
Code to Load Instance:

instance = load_instance('problem_9_4a')

Code to Build Manually:

instance = {
    "holding_cost": 275,
    "fixed_cost": 2250,
    "demand_rate": 182500,
    "yield_mean": -50.0,
    "yield_sd": 0.02
}

Problem 9.4(b) (EOQ with multiplicative yield uncertainty)

Name: problem_9_4b
Description: Problem 9.4(b) (EOQ with multiplicative yield uncertainty)
Code to Load Instance:

instance = load_instance('problem_9_4b')

Code to Build Manually:

instance = {
    "holding_cost": 275,
    "fixed_cost": 2250,
    "demand_rate": 182500,
    "yield_mean": 0.9,
    "yield_sd": 0.05773502691896258
}

Example 9.6 (newsvendor with additive yield uncertainty)

Name: example_9_6
Description: Example 9.6 (newsvendor with additive yield uncertainty)
Code to Load Instance:

instance = load_instance('example_9_6')

Code to Build Manually:

instance = {
    "holding_cost": 15000000,
    "stockout_cost": 75000000,
    "demand": 1.5,
    "yield_lo": -0.5,
    "yield_hi": 0.5
}

Problem 9.5 (newsvendor with additive yield uncertainty)

Name: problem_9_5
Description: Problem 9.5 (newsvendor with additive yield uncertainty)
Code to Load Instance:

instance = load_instance('problem_9_5')

Code to Build Manually:

instance = {
    "holding_cost": 150,
    "stockout_cost": 1200,
    "demand": 25,
    "yield_lo": -5,
    "yield_hi": 0
}

API Reference¶

load_instance(instance_name, filepath=None, ignore_state_vars=True)[source]¶

Load an instance from a JSON file.

If the instance was originally specified as a SupplyChainNetwork object, returns the object; otherwise, returns the instance in a dictionary.

Parameters

instance_name (str) – The name of the instance.
filepath (str, optional) – Path to the JSON file. If None, the function determines the path to datasets/stockpyl_instances.json.
ignore_state_vars (bool, optional) – If True, function will ignore any saved state variables in the nodes.

Returns

The loaded instance. If the instance was originally specified as a SupplyChainNetwork object, returns the object; otherwise, returns the instance in a dictionary in which the keys equal the parameter names (e.g., “holding_cost”) and the values equal the parameter values (e.g., 0.5).

Return type

dict or SupplyChainNetwork

Raises

ValueError – If the JSON file does not exist or the instance cannot be found in the JSON file.

save_instance(instance_name, instance_data, instance_description='', filepath=None, replace=True, create_if_none=True, delete_if_exists=False, omit_state_vars=True)[source]¶

Save an instance to a JSON file.

Parameters

instance_name (str) – The name of the instance. This will be used later for retreving the instance.
instance_data (dict or SupplyChainNetwork) – The instance data as a dictionary (with keys equal to parameter names (e.g., “holding_cost”) and values equal to parameter values (e.g., 0.5)) or as a SupplyChainNetwork object (in which case the instance is serialized using jsonpickle).
instance_description (str, optional) – A longer descrtiption of the instance.
filepath (str, optional) – Path to the JSON file. If None, the function determines the path to datasets/stockpyl_instances.json.
replace (bool, optional) – If an instance with the same instance_name is already in the file, the function will replace it if True and will ignore it (and write nothing) if False.
create_if_none (bool, optional) – If the file does not already exist, the function will create a new file if True; otherwise, it will not do anything and issue a warning.
delete_if_exists (bool, optional) – If the file already exists, the function will delete it first if True; otherwise, it will modify the existing file.
omit_state_vars (bool, optional) – If True, the function will not save state variables as part of the nodes, even if they are present in the instance.

instances Module¶

Overview¶

API Reference¶

`instances` Module¶