Price list behind API: Difference between revisions

From Rosetta Code
Content added Content deleted
m (→‎{{header|Phix}}: default comment)
(Added Wren)
Line 161: Line 161:
From 90215.0 ... 95249.0 with 4999 items.
From 90215.0 ... 95249.0 with 4999 items.
From 95250.0 ... 104742.0 with 4864 items.</pre>
From 95250.0 ... 104742.0 with 4864 items.</pre>

=={{header|Wren}}==
{{trans|Python}}
{{libheader|Wren-math}}
{{libheader|Wren-fmt}}
<lang ecmascript>import "random" for Random
import "/math" for Nums
import "/fmt" for Fmt

var rand = Random.new()

var getMaxPrice = Fn.new { |prices| Nums.max(prices) }

var getPrangeCount = Fn.new { |prices, min, max| prices.count { |p| p >= min && p <= max } }

var get5000 = Fn.new { |prices, min, max, n|
var count = getPrangeCount.call(prices, min, max)
var delta = ((max - min)/2).floor
while (count != n && delta > 0) {
max = (count > n) ? max-delta : max+delta
count = getPrangeCount.call(prices, min, max)
delta = (delta/2).floor
}
return [max, count]
}

var getAll5000 = Fn.new { |prices, min, max, n|
var mc = get5000.call(prices, min, max, n)
var pmax = mc[0]
var pcount = mc[1]
var res = [[min, pmax, pcount]]
while (pmax < max) {
var pmin = pmax + 1
mc = get5000.call(prices, pmin, max, n)
pmax = mc[0]
pcount = mc[1]
res.add([pmin, pmax, pcount])
}
return res
}
var numPrices = 1e5
var maxPrice = 1e5
var prices = List.filled(numPrices, 0) // list of prices
for (i in 1..numPrices) prices[i-1] = rand.int(maxPrice + 1)
var actualMax = getMaxPrice.call(prices)
System.print("Using %(numPrices) items with prices from 0 to %(actualMax):")
var res = getAll5000.call(prices, 0, actualMax, 5000)
System.print("Split into %(res.count) bins of approx 5000 elements:")
var total = 0
for (r in res) {
var min = r[0]
var max = r[1]
if (max > actualMax) max = actualMax
var cnt = r[2]
total = total + cnt
Fmt.print(" From $6d to $6d with $4d items", min, max, cnt)
}
if (total != numPrices) {
System.print("Something went wrong - grand total of %(total) doesn't equal %(numPrices)!")
}</lang>

{{out}}
Sample run:
<pre>
Using 100000 items with prices from 0 to 99998:
Split into 20 bins of approx 5000 elements:
From 0 to 5043 with 5000 items
From 5044 to 10102 with 5001 items
From 10103 to 15192 with 5000 items
From 15193 to 20320 with 5000 items
From 20321 to 25368 with 4998 items
From 25369 to 30376 with 5003 items
From 30377 to 35422 with 5001 items
From 35423 to 40337 with 5001 items
From 40338 to 45299 with 5000 items
From 45300 to 50389 with 5001 items
From 50390 to 55402 with 4998 items
From 55403 to 60382 with 5001 items
From 60383 to 65330 with 4999 items
From 65331 to 70278 with 5001 items
From 70279 to 75285 with 4999 items
From 75286 to 80336 with 5000 items
From 80337 to 85289 with 5000 items
From 85290 to 90291 with 5000 items
From 90292 to 95052 with 5001 items
From 95053 to 99998 with 4996 items
</pre>

Revision as of 15:29, 25 November 2020

Price list behind API is a draft programming task. It is not yet considered ready to be promoted as a complete task, for reasons that should be found in its talk page.

There is a list of around 100_000 prices in the range £0 to £100_000, expressed in whole £, (no pence); and prices may be duplicated.
The API allows access to the maximum item price via function get_max_price(); and the number of items equal-to and between two given price points via function get_prange_count(pricemin, pricemax).
Assume that for the purposes of testing, you have access to the actual number of priced items to split.

Task
  1. Write functions to randomly generate around 100K prices and provide the get_prange_count and get_max_price API calls.
  2. Write functions to provide non-overlapping min and max price ranges that provide product counts where most are close to, but no more than, 5_000.
  3. Ensure that all priced items are covered by all the ranges of prices shown
  4. Show ascending price ranges and the number of items covered by each range.
  5. Show output from a sample run here.

Phix

Translation of: Python

Note that defaulted arguments of the form mx=get_max_price() are not currently supported, hence a slightly hacky workaround.
If you defined constant mp = get_max_price(), then mx=mp style parameter defaulting would be fine. <lang Phix>constant price_list_size = 99_000 + rand(2_001) - 1,

        price_list = sq_sub(sq_rand(repeat(100_000,price_list_size)),1),
        delta_price = 1 -- Minimum difference between any two different prices.

function get_prange_count(integer startp, endp)

   return length(filter(price_list,"in",{startp,endp},"[]"))

end function

function get_max_price()

   return max(price_list)

end function

function get_5k(integer mn=0, mx=-1, num=5_000)

   if mx=-1 then mx = get_max_price() end if
   -- Binary search for num items between mn and mx, adjusting mx
   integer count = get_prange_count(mn, mx)
   atom delta_mx = (mx - mn) / 2
   while count != num and delta_mx >= delta_price / 2 do
       mx = floor(mx + iff(count > num ? -delta_mx : +delta_mx))
       {count, delta_mx} = {get_prange_count(mn, mx), delta_mx / 2}
   end while
   return {mx, count}

end function

function get_all_5k(integer mn=0, mx=-1, num=5_000)

   if mx=-1 then mx = get_max_price() end if
   -- Get all non-overlapping ranges
   integer {partmax, partcount} = get_5k(mn, mx, num)
   sequence result = Template:Mn, partmax, partcount
   while partmax < mx do
       integer partmin = partmax + delta_price 
       {partmax, partcount} = get_5k(partmin, mx, num)
       result = append(result,{partmin, partmax, partcount})
   end while
   return result

end function

printf(1,"Using %d random prices from 0 to %d\n",{price_list_size,get_max_price()}) sequence result = get_all_5k() printf(1,"Splits into %d bins of approx 5000 elements\n",{length(result)}) for i=1 to length(result) do

   printf(1,"  From %8.1f ... %8.1f with %d items.\n",result[i])

end for if length(price_list) != sum(vslice(result,3)) then

   printf(1,"\nWhoops! Some items missing:\n")

end if</lang>

Output:
Using 99714 random prices from 0 to 99999
Splits into 20 bins of approx 5000 elements
  From      0.0 ...   4977.0 with 5000 items.
  From   4978.0 ...  10019.0 with 4999 items.
  From  10020.0 ...  15114.0 with 4999 items.
  From  15115.0 ...  19987.0 with 4998 items.
  From  19988.0 ...  25088.0 with 4996 items.
  From  25089.0 ...  30080.0 with 4995 items.
  From  30081.0 ...  35117.0 with 5000 items.
  From  35118.0 ...  40081.0 with 4999 items.
  From  40082.0 ...  45080.0 with 5000 items.
  From  45081.0 ...  50181.0 with 5000 items.
  From  50182.0 ...  55223.0 with 5000 items.
  From  55224.0 ...  60271.0 with 5000 items.
  From  60272.0 ...  65102.0 with 4999 items.
  From  65103.0 ...  70140.0 with 5000 items.
  From  70141.0 ...  75195.0 with 4997 items.
  From  75196.0 ...  80203.0 with 4998 items.
  From  80204.0 ...  85210.0 with 4999 items.
  From  85211.0 ...  90182.0 with 5000 items.
  From  90183.0 ...  95268.0 with 4999 items.
  From  95269.0 ... 104722.0 with 4736 items.

Python

<lang python>import random

  1. %%Sample price generation

price_list_size = random.choice(range(99_000, 101_000)) price_list = random.choices(range(100_000), k=price_list_size)

delta_price = 1 # Minimum difference between any two different prices.

  1. %% API

def get_prange_count(startp, endp):

   return len([r for r in price_list if startp <= r <= endp])

def get_max_price():

   return max(price_list)
  1. %% Solution

def get_5k(mn=0, mx=get_max_price(), num=5_000):

   "Binary search for num items between mn and mx, adjusting mx"
   count = get_prange_count(mn, mx)
   delta_mx = (mx - mn) / 2
   while count != num and delta_mx >= delta_price / 2:
       mx += -delta_mx if count > num else +delta_mx
       mx = mx // 1    # Floor
       count, delta_mx = get_prange_count(mn, mx), delta_mx / 2
   return mx, count

def get_all_5k(mn=0, mx=get_max_price(), num=5_000):

   "Get all non-overlapping ranges"
   partmax, partcount = get_5k(mn, mx, num)
   result = [(mn, partmax, partcount)]
   while partmax < mx:
       partmin = partmax + delta_price 
       partmax, partcount = get_5k(partmin, mx, num)
       result.append((partmin, partmax, partcount))
   return result

if __name__ == '__main__':

   print(f"Using {price_list_size} random prices from 0 to {get_max_price()}")
   result = get_all_5k()
   print(f"Splits into {len(result)} bins of approx 5000 elements")
   for mn, mx, count in result:
       print(f"  From {mn:8.1f} ... {mx:8.1f} with {count} items.")
   if len(price_list) != sum(count for mn, mx, count in result):
       print("\nWhoops! Some items missing:")</lang>
Output:
Using 99838 random prices from 0 to 99999
Splits into 20 bins of approx 5000 elements
  From      0.0 ...   4876.0 with 4999 items.
  From   4877.0 ...   9973.0 with 4997 items.
  From   9974.0 ...  14954.0 with 4999 items.
  From  14955.0 ...  20041.0 with 4997 items.
  From  20042.0 ...  25132.0 with 4999 items.
  From  25133.0 ...  30221.0 with 5000 items.
  From  30222.0 ...  35313.0 with 5000 items.
  From  35314.0 ...  40263.0 with 5000 items.
  From  40264.0 ...  45249.0 with 4997 items.
  From  45250.0 ...  50264.0 with 5000 items.
  From  50265.0 ...  55251.0 with 5000 items.
  From  55252.0 ...  60301.0 with 4997 items.
  From  60302.0 ...  65239.0 with 5000 items.
  From  65240.0 ...  70220.0 with 4998 items.
  From  70221.0 ...  75193.0 with 4999 items.
  From  75194.0 ...  80229.0 with 4996 items.
  From  80230.0 ...  85191.0 with 4997 items.
  From  85192.0 ...  90214.0 with 5000 items.
  From  90215.0 ...  95249.0 with 4999 items.
  From  95250.0 ... 104742.0 with 4864 items.

Wren

Translation of: Python
Library: Wren-math
Library: Wren-fmt

<lang ecmascript>import "random" for Random import "/math" for Nums import "/fmt" for Fmt

var rand = Random.new()

var getMaxPrice = Fn.new { |prices| Nums.max(prices) }

var getPrangeCount = Fn.new { |prices, min, max| prices.count { |p| p >= min && p <= max } }

var get5000 = Fn.new { |prices, min, max, n|

   var count = getPrangeCount.call(prices, min, max)
   var delta = ((max - min)/2).floor
   while (count != n && delta > 0) {
       max = (count > n) ? max-delta : max+delta
       count = getPrangeCount.call(prices, min, max)
       delta = (delta/2).floor
   }
   return [max, count]

}

var getAll5000 = Fn.new { |prices, min, max, n|

   var mc = get5000.call(prices, min, max, n)
   var pmax = mc[0]
   var pcount = mc[1]
   var res = min, pmax, pcount
   while (pmax < max) {
       var pmin = pmax + 1
       mc = get5000.call(prices, pmin, max, n)
       pmax = mc[0]
       pcount = mc[1]
       res.add([pmin, pmax, pcount])
   }
   return res

} var numPrices = 1e5 var maxPrice = 1e5 var prices = List.filled(numPrices, 0) // list of prices for (i in 1..numPrices) prices[i-1] = rand.int(maxPrice + 1) var actualMax = getMaxPrice.call(prices) System.print("Using %(numPrices) items with prices from 0 to %(actualMax):") var res = getAll5000.call(prices, 0, actualMax, 5000) System.print("Split into %(res.count) bins of approx 5000 elements:") var total = 0 for (r in res) {

   var min = r[0]
   var max = r[1]
   if (max > actualMax) max = actualMax
   var cnt = r[2]
   total = total + cnt
   Fmt.print("   From $6d to $6d with $4d items", min, max, cnt)

} if (total != numPrices) {

   System.print("Something went wrong - grand total of %(total) doesn't equal %(numPrices)!")

}</lang>

Output:

Sample run:

Using 100000 items with prices from 0 to 99998:
Split into 20 bins of approx 5000 elements:
   From      0 to   5043 with 5000 items
   From   5044 to  10102 with 5001 items
   From  10103 to  15192 with 5000 items
   From  15193 to  20320 with 5000 items
   From  20321 to  25368 with 4998 items
   From  25369 to  30376 with 5003 items
   From  30377 to  35422 with 5001 items
   From  35423 to  40337 with 5001 items
   From  40338 to  45299 with 5000 items
   From  45300 to  50389 with 5001 items
   From  50390 to  55402 with 4998 items
   From  55403 to  60382 with 5001 items
   From  60383 to  65330 with 4999 items
   From  65331 to  70278 with 5001 items
   From  70279 to  75285 with 4999 items
   From  75286 to  80336 with 5000 items
   From  80337 to  85289 with 5000 items
   From  85290 to  90291 with 5000 items
   From  90292 to  95052 with 5001 items
   From  95053 to  99998 with 4996 items