demand_source Module

Overview

This module contains the DemandSource class. A DemandSource object represents external demand observed by a node. The demand can be random or deterministic. Attributes specify the type of demand distribution and its parameters. The object can generate demands from the specified distribution.

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).

Example: Create a DemandSource object representing demand that has a normal distribution with a mean of 50 and a standard deviation of 10. Generate a random demand from the distribution.

>>> ds = DemandSource(type='N', mean=50, standard_deviation=10)
>>> ds.generate_demand()        
46.75370030596123
>>> # Tell object to round demands to integers.
>>> ds.round_to_int = True
>>> ds.generate_demand()        
63

API Reference

class DemandSource(**kwargs)

A DemandSource object represents external demand observed by a node. The demand can be random or deterministic. Attributes specify the type of demand distribution and its parameters. The object can generate demands from the specified distribution.

Parameters

**kwargs – Keyword arguments specifying values of one or more attributes of the DemandSource, e.g., type='N'.

type

The demand type, as a string. Currently supported strings are:

• None

• ‘N’ (normal)

• ‘P’ (Poisson)

• ‘UD’ (uniform discrete)

• ‘UC’ (uniform continuous)

• ‘NB’ (negative binomial)

• ‘D’ (deterministic)

• ‘CD’ (custom discrete)

Type

str

round_to_int

Round demand to nearest integer?

Type

bool

demand_list

List of demands, one per period (for deterministic demand types), or list of possible demand values (for custom discrete demand types). For deterministic demand types, if demand is required in a period beyond the length of the list, the list is restarted at the beginning. This also allows demand_list to be a singleton, in which case it is used in every period. Required if type == ‘D’ or ‘CD’. [\(d\)]

Type

list, optional

probabilities

List of probabilities of each demand value (for custom discrete demand types). Required if type == ‘CD’.

Type

list, optional

lo

Low value of demand range (for uniform demand types). Required if type == ‘UD’ or ‘UC’.

Type

float, optional

hi

High value of demand range (for uniform demand types). Required if type == ‘UD’ or ‘UC’.

Type

float, optional

n

Parameter for negative binomial distribution indicating number of trial successes. Required if type == ‘NB’. [\(n\)]

Type

int, optional

p

Parameter for negative binomial distribution indicating probability of success for each trial. Required if type == ‘NB’. [\(p\)]

Type

float, optional

property mean

Return mean set by user, if any; or, for distributions whose mean is not set but is calculated from other parameters, returns the calculated mean. If neither is true, return None.

property standard_deviation

Return standard deviation set by user, if any; or, for distributions whose standard deviation is not set but is calculated from other parameters, returns the calculated standard deviation. If neither is true, return None.

property demand_distribution

Demand distribution, as a scipy.stats.rv_continuous or scipy.stats.rv_discrete object. Returns None if demand source type is 'D'. Read only.

property is_discrete

True if the distribution is discrete, False if it is continuous. Read only.

The distribution is discrete if self.type is ‘P’, ‘UD’, ‘CD’, ‘NB’, or ‘D’.

Returns

True if the distribution is discrete, False if it is continuous.

Return type

bool

initialize()

Initialize the parameters in the object to their default values.

validate_parameters()

Check that appropriate parameters have been provided for the given demand type. Raise an exception if not.

to_dict()

Convert the DemandSource object to a dict. List attributes (demand_list, probabilities) are deep-copied so changes to the original object do not get propagated to the dict.

Returns

The dict representation of the object.

Return type

dict

classmethod from_dict(the_dict)

Return a new DemandSource object with attributes copied from the values in the_dict. List attributes (demand_list, probabilities) are deep-copied so changes to the original dict do not get propagated to the object. Any missing attributes are set to their default values.

Parameters

the_dict (dict) – Dict representation of a DemandSource, typically created using to_dict().

Returns

The object converted from the dict.

Return type

DemandSource

generate_demand(period=None)

Generate a demand value using the demand type specified in type. If type is None, returns None.

Parameters

period (int, optional) – The period to generate a demand value for. If type = ‘D’ (deterministic), this is required if demand_list is a list of demands, one per period. If omitted, will return first (or only) demand in list.

Returns

demand – The demand value.

Return type

float

cdf(x)

Cumulative distribution function of demand distribution.

In some cases, this is just a wrapper around cdf() function of scipy.stats.rv_continuous or scipy.stats.rv_discrete object.

Parameters

x (float) – Value to calculate cdf for.

Returns

F – cdf of x.

Return type

float

lead_time_demand_distribution(lead_time)

Return lead-time demand distribution, as a scipy.stats.rv_continuous or scipy.stats.rv_discrete object.

Note

If lead_time equals 0, this method returns an rv_discrete object with a single value (0) and a single probability (1).

Note

For ‘UC’, ‘UD’, ‘NB’, and ‘CD’ demands, this method calculates the lead-time demand distribution as the sum of lead_time independent random variables. Therefore, the method requires lead_time to be an integer for these distributions. If it is not, it raises a ValueError.

Parameters

lead_time (float or int) – The lead time. [\(L\)]

Returns

distribution – The lead-time demand distribution object.

Return type

rv_continuous or rv_discrete

Raises

ValueError – If type is ‘UC’, ‘UD’, ‘NB’, or ‘CD’ and lead_time is not an integer.