fiscal-calendar-helper 1.2.4

A fiscal helper for retail calendars

The Fiscal Calendar Helper is a set of fiscal calendar helper functions that enables easy use and navigation of a fiscal calendar in Python 3 projects.

API Reference:

• Installation
• The Fiscal Calendar
• Time Offsets - Moving between time periods
  • fiscal.offset()
  • fiscal.next()
  • fiscal.previous()
• Time Conversions - Converting between time levels
  • fiscal.convert()
• Parent / Child Time Relationships - Managing time in with the hierarchy
  • fiscal.fill_to()
  • fiscal.unpack()
• More Info - Repo, bugs, license info

Installation

To use the fiscal helper:

from fiscal_calendar_helper import fiscal 

The Fiscal Calendar

The current fiscal calendar is based on the National Retail Federation's 4-5-4 calendar. In the future, we plan to support other calendars (Gregorian), different starting days of week, and longer time horizons.

The current supported years are 2016 - 2025

Time Levels

Our fiscal calendar is capable of supportting different levels of the calendar hierarchy. While many of the examples may look similar to the other levels, their structures are not interchangable and mean something specific to that level.

Level	Description	Example	Data Type	Relative or Absolute
key	This is a numeric key for this record and should not be confused with the date itself. This is most likely not what you're looking for. It's a structure holding counts of year, week, and day_of_week as an index starting at 1. For example this key 2023_07_7 actually represents the date 3/18/23 which looks totally different.	2023_07_7	string	Absolute
week_id	This is the unique id for this fiscal week	2023_07	string	Absolute
year	This is the fiscal year.	2023	string	Absolute
quarter	This is the fiscal quarter.	Q1	string	Relative
month	This is the fiscal month.	March	string	Relative
month_id	This is the unique id for the fiscal month.	2023_M02	string	Absolute
week_num	This is a relative count of the number of weeks in the year.	7	string	Relative
day_of_week	This is the day of the week in plain text.	Saturday	string	Absolute
date	This is a datetime formatted representation of the fiscal date.	2023-03-18	datetime	Absolute
date_id	This is the id for this date.	2023_03_18	string	Absolute

Relative vs Absolute Levels

Absolute levels mean that this column only uses one value for a specific time period, where as relative levels mean that the value in the column may be repeated over several time periods.

A week_id such as 2023_07 is absolute because it will only occur once on the calendar, where as week_num such as 7 because that will exist for every year in the calendar.

Most of the time, you'll likely be working with absolute dates only - like finding the starting date for a particular week. But you may also want to compare months (Relative) across several years (Absolute) to create a table where values can be compared directly.

fiscal.calendar

The actual fiscal calendar is stored as a pandas table located at fiscal.calendar.

Time Offsets

All of the time offsets are currently only built to work with week_id or date_id levels.

fiscal.offset()

offset(input, offset)

Returns week_id or date_id offset by specified periods.

>>> fiscal.offset('2022_31', 4)
'2022_35'

fiscal.next()

next(input)

Returns the next week or date.

>>> fiscal.next('2022_31')
'2022_32'

fiscal.previous()

previous(input)
Returns the previous week or date.

>>> fiscal.previous('2022_31')
'2022_30'

Time Conversions

Time conversions allow for easy translation between years, weeks, week nums, etc without having to parse strings.

fiscal.convert()

fiscal.convert(input, inp_level, out_level)

The convert function takes an input and level type, and output level and converts the value. If there are multiple valid answers, it will return the first time in the list.

Example: Find the starting date for a fiscal week

>>> fiscal.convert('2022_36', 'week_id', 'date')
['2016_02_14', '2016_02_15', '2016_02_16', '2016_02_17']

Parent / Child Time Relationships

Time is structured into a hierarchy, which means that there are parent and child elements in the hierarchy structure. You can read more on parent-child hierarchies if the concept is unfamiliar.

For our purposes a parent-like level will be any level that contains multiple child elements of the same type.

Examples:

• year is parent-like to quarter
• week is child-like to quarter
• week is parent-like to date and date_id
• year is parent-like to date
• date is child-like to month

fiscal.fill_to()

fill_to(input, child_level = 'date_id', parent_level = 'week_id', direction = 'start')

The fill_to function takes a child input as a starting point, and then a parent output as a container, and lists the remaining children in that parent.

Example: Given a date, find the remaining dates in the week.

>>> fiscal.fill_to('2016_02_18', 'date_id', 'week_id', 'start')
['2016_02_14', '2016_02_15', '2016_02_16', '2016_02_17']

fiscal.unpack()

unpack(input, parent_level, child_level)

The unpack function takes in a parent level and returns all the children contained withint it.

Example: Return all dates within a quarter.

>>> fiscal.unpack('2016_01', 'week_id', 'date_id')
['2016_02_07', '2016_02_08', '2016_02_09', '2016_02_10', '2016_02_11', '2016_02_12', '2016_02_13']

More info

• Author: Phillip Geltman
• License Type: MIT License
• Github Repo: https://github.com/Growthsayer/fiscal-helper
• Bug Tracker: https://github.com/Growthsayer/fiscal-helper/issues