# Example 905 — The factory saw a season the customers did not have `the_factory_saw_a_season_the_customers_did_not_have.eml` - A factory plans capacity from the orders it receives, the orders are real and exactly recorded, and they show a sharp spike on the first of every month. Where that spike comes from is computed below. ## EML ```eml # Self-authored for the EML case corpus (no external origin). A factory plans # capacity from the orders it receives, the orders are real and exactly # recorded, and they show a sharp spike on the first of every month. Where that # spike comes from is computed below. # # The planning is careful. It reads the real order book, not a forecast; every # order is dated to the day it arrived; the monthly spike is genuinely there; # and the intent is exactly 'staff up for the busy period'. # # Customers buy a flat hundred a day, and the retailer orders once a month in # one batch, so the factory sees thirty days of demand land on one day and a # calm that is not calm on the other twenty-nine. 100 => customer_demand_per_day 30 => days_per_month customer_demand_per_day * days_per_month => retailer_monthly_batch retailer_monthly_batch => factory_orders_on_the_first 0 => factory_orders_on_other_days customer_demand_per_day => customer_demand_on_the_first customer_demand_per_day => customer_demand_on_other_days int(factory_orders_on_the_first / customer_demand_per_day) => apparent_peak_as_a_multiple_of_daily_demand customer_demand_on_the_first - customer_demand_on_other_days => customers_actual_swing factory_orders_on_the_first - factory_orders_on_other_days => factorys_apparent_swing "customer demand, every day : " + str(customer_demand_per_day) ^0 "retailer's batch, once a month : " + str(retailer_monthly_batch) ^0 "" ^0 "factory orders, the first : " + str(factory_orders_on_the_first) ^0 "factory orders, other days : " + str(factory_orders_on_other_days) ^0 "customer demand, the first : " + str(customer_demand_on_the_first) ^0 "customer demand, other days : " + str(customer_demand_on_other_days) ^0 "" ^0 "apparent peak : " + str(apparent_peak_as_a_multiple_of_daily_demand) + " times a normal day" ^0 "customers' swing, day to day : " + str(customers_actual_swing) ^0 "factory's apparent swing : " + str(factorys_apparent_swing) ^0 "" ^0 # ---- what the planning verified ---- "the capacity plan" ^0 " reads : the real order book, not a forecast" ^0 " dating : every order to the day it arrived" ^0 " the spike : genuinely in the data, every month" ^0 " intent : staff up for the busy period" ^0 " orders misdated : 0" ^0 " verdict : DEMAND PEAKS THIRTY-FOLD ON THE FIRST" ^0 "" ^0 " reading the real order book with real dates is the part" ^0 " done right here, and it is why the spike is not an" ^0 " artefact of the factory's own records" ^0 "" ^0 # ---- where the spike comes from ---- "batching" ^0 " what customers do : buy " + str(customer_demand_per_day) + " a day, every day" ^0 " what the retailer does : orders " + str(days_per_month) + " days at once, on the first" ^0 " what the factory sees : " + str(retailer_monthly_batch) + " on one day, nothing on the rest" ^0 " so the season : is in the retailer's calendar, not the" ^0 " customers' behaviour" ^0 " what staffing up for the first buys : capacity for a" ^0 " peak that is an accounting event" ^0 "" ^0 # ---- what the factory got ---- "the plan" ^0 " temporary staff hired for : the first of each month" ^0 " demand on the first, from customers : " + str(customer_demand_on_the_first) + ", same as any day" ^0 " is the order book wrong : no; the batch really arrives" ^0 " then" ^0 " is an order date a demand date : no; it is the date the" ^0 " retailer chose to send thirty days at once" ^0 "" ^0 # ---- null control ---- # The same factory planning on the retailer's sell-through (customer sales per # day) instead of on the retailer's order dates. 3000 => nc_peak_planned_for_on_order_dates 100 => nc_peak_planned_for_on_sell_through 0 => nc_seasonal_staff_needed_on_sell_through "null control - plan on sell-through, not on order dates" ^0 " peak planned for, order dates : " + str(nc_peak_planned_for_on_order_dates) ^0 " peak planned for, sell-through : " + str(nc_peak_planned_for_on_sell_through) ^0 " seasonal staff needed on sell-through : " + str(nc_seasonal_staff_needed_on_sell_through) ^0 " no customer and no order changed; the calendar the" ^0 " factory read stopped being the retailer's" ^0 "" ^0 # ---- the rule ---- "what an order-book capacity plan guarantees" ^0 " the plan matches when orders arrive : exactly, real" ^0 " orders, real dates" ^0 " the plan matches when customers buy : not addressed; the" ^0 " retailer batches " + str(days_per_month) + " days into one order, so the factory" ^0 " sees a " + str(apparent_peak_as_a_multiple_of_daily_demand) + "-fold peak in a demand that swings by " + str(customers_actual_swing) ^0 "" ^0 "the rhythm of the orders is the rhythm of whoever places them, and one tier's" ^0 "convenience arrives at the next as weather; a spike that is really a batch has" ^0 "a date but no cause the customers would recognise" ^0 "" ^0 "It plans from the real order book with real dates - the monthly spike is truly" ^0 "there. But customers buy " + str(customer_demand_per_day) + " every day and the retailer batches " + str(days_per_month) + " days into" ^0 "one order, so the factory staffs for a " + str(apparent_peak_as_a_multiple_of_daily_demand) + "-fold peak in a demand that" ^0 "swings by " + str(customers_actual_swing) + ", until it plans on sell-through instead of order dates." ^0 ``` ## Python (deterministic transpilation) ```python customer_demand_per_day = 100 days_per_month = 30 retailer_monthly_batch = customer_demand_per_day * days_per_month factory_orders_on_the_first = retailer_monthly_batch factory_orders_on_other_days = 0 customer_demand_on_the_first = customer_demand_per_day customer_demand_on_other_days = customer_demand_per_day apparent_peak_as_a_multiple_of_daily_demand = int(factory_orders_on_the_first / customer_demand_per_day) customers_actual_swing = customer_demand_on_the_first - customer_demand_on_other_days factorys_apparent_swing = factory_orders_on_the_first - factory_orders_on_other_days print("customer demand, every day : " + str(customer_demand_per_day)) print("retailer's batch, once a month : " + str(retailer_monthly_batch)) print("") print("factory orders, the first : " + str(factory_orders_on_the_first)) print("factory orders, other days : " + str(factory_orders_on_other_days)) print("customer demand, the first : " + str(customer_demand_on_the_first)) print("customer demand, other days : " + str(customer_demand_on_other_days)) print("") print("apparent peak : " + str(apparent_peak_as_a_multiple_of_daily_demand) + " times a normal day") print("customers' swing, day to day : " + str(customers_actual_swing)) print("factory's apparent swing : " + str(factorys_apparent_swing)) print("") print("the capacity plan") print(" reads : the real order book, not a forecast") print(" dating : every order to the day it arrived") print(" the spike : genuinely in the data, every month") print(" intent : staff up for the busy period") print(" orders misdated : 0") print(" verdict : DEMAND PEAKS THIRTY-FOLD ON THE FIRST") print("") print(" reading the real order book with real dates is the part") print(" done right here, and it is why the spike is not an") print(" artefact of the factory's own records") print("") print("batching") print(" what customers do : buy " + str(customer_demand_per_day) + " a day, every day") print(" what the retailer does : orders " + str(days_per_month) + " days at once, on the first") print(" what the factory sees : " + str(retailer_monthly_batch) + " on one day, nothing on the rest") print(" so the season : is in the retailer's calendar, not the") print(" customers' behaviour") print(" what staffing up for the first buys : capacity for a") print(" peak that is an accounting event") print("") print("the plan") print(" temporary staff hired for : the first of each month") print(" demand on the first, from customers : " + str(customer_demand_on_the_first) + ", same as any day") print(" is the order book wrong : no; the batch really arrives") print(" then") print(" is an order date a demand date : no; it is the date the") print(" retailer chose to send thirty days at once") print("") nc_peak_planned_for_on_order_dates = 3000 nc_peak_planned_for_on_sell_through = 100 nc_seasonal_staff_needed_on_sell_through = 0 print("null control - plan on sell-through, not on order dates") print(" peak planned for, order dates : " + str(nc_peak_planned_for_on_order_dates)) print(" peak planned for, sell-through : " + str(nc_peak_planned_for_on_sell_through)) print(" seasonal staff needed on sell-through : " + str(nc_seasonal_staff_needed_on_sell_through)) print(" no customer and no order changed; the calendar the") print(" factory read stopped being the retailer's") print("") print("what an order-book capacity plan guarantees") print(" the plan matches when orders arrive : exactly, real") print(" orders, real dates") print(" the plan matches when customers buy : not addressed; the") print(" retailer batches " + str(days_per_month) + " days into one order, so the factory") print(" sees a " + str(apparent_peak_as_a_multiple_of_daily_demand) + "-fold peak in a demand that swings by " + str(customers_actual_swing)) print("") print("the rhythm of the orders is the rhythm of whoever places them, and one tier's") print("convenience arrives at the next as weather; a spike that is really a batch has") print("a date but no cause the customers would recognise") print("") print("It plans from the real order book with real dates - the monthly spike is truly") print("there. But customers buy " + str(customer_demand_per_day) + " every day and the retailer batches " + str(days_per_month) + " days into") print("one order, so the factory staffs for a " + str(apparent_peak_as_a_multiple_of_daily_demand) + "-fold peak in a demand that") print("swings by " + str(customers_actual_swing) + ", until it plans on sell-through instead of order dates.") ``` ## stdout (executed) ```text customer demand, every day : 100 retailer's batch, once a month : 3000 factory orders, the first : 3000 factory orders, other days : 0 customer demand, the first : 100 customer demand, other days : 100 apparent peak : 30 times a normal day customers' swing, day to day : 0 factory's apparent swing : 3000 the capacity plan reads : the real order book, not a forecast dating : every order to the day it arrived the spike : genuinely in the data, every month intent : staff up for the busy period orders misdated : 0 verdict : DEMAND PEAKS THIRTY-FOLD ON THE FIRST reading the real order book with real dates is the part done right here, and it is why the spike is not an artefact of the factory's own records batching what customers do : buy 100 a day, every day what the retailer does : orders 30 days at once, on the first what the factory sees : 3000 on one day, nothing on the rest so the season : is in the retailer's calendar, not the customers' behaviour what staffing up for the first buys : capacity for a peak that is an accounting event the plan temporary staff hired for : the first of each month demand on the first, from customers : 100, same as any day is the order book wrong : no; the batch really arrives then is an order date a demand date : no; it is the date the retailer chose to send thirty days at once null control - plan on sell-through, not on order dates peak planned for, order dates : 3000 peak planned for, sell-through : 100 seasonal staff needed on sell-through : 0 no customer and no order changed; the calendar the factory read stopped being the retailer's what an order-book capacity plan guarantees the plan matches when orders arrive : exactly, real orders, real dates the plan matches when customers buy : not addressed; the retailer batches 30 days into one order, so the factory sees a 30-fold peak in a demand that swings by 0 the rhythm of the orders is the rhythm of whoever places them, and one tier's convenience arrives at the next as weather; a spike that is really a batch has a date but no cause the customers would recognise It plans from the real order book with real dates - the monthly spike is truly there. But customers buy 100 every day and the retailer batches 30 days into one order, so the factory staffs for a 30-fold peak in a demand that swings by 0, until it plans on sell-through instead of order dates. ``` ## Round-trip `ok: true` — round-trip fixpoint reached (python1 == python2) ## Trace event types eml:run:start · eml:assign · eml:output · eml:run:done