NEW: Sort by Face Value

[Dana-Levitas]

Following an initiative by collectors a new feature has been added to the Colnect catalog - Sort by Face Value. All collectible categories that display Face Value now have the option to view item list pages sorted by the face value of the items. See for example this Stamps list page.

sort by face value.png

Currently on this sort option:

- All non-numeric values appear at the end.

- The bigger values are at the top.

- Same face values are sorted by catalog code, then issue date, then name.

You are welcome to try is out and let us know what you think, is it comfortable for you?

[Luis93]

I find it very useful.

But in the future if possible could there be an option to choose the numeric order, for example from smaller to bigger values or as we have now from bigger to smaller values.

Thanks

[footy68]

Descending order makes NO sense. Who is sorting the stamps like this?

Otherwise: 1 Dirham is higher than 500 Fils. How is this solved in future?

For me is it is not useful, sorry. But I am only one who voted for 'Not useful'.

[tcawe]

But in the future if possible could there be an option to choose the numeric order, for example from smaller to bigger values or as we have now from bigger to smaller values.

Descending order makes NO sense. Who is sorting the stamps like this?

We can shift to ascending order but what should we do then with stamps that have no face value or a face value that's not a number?

Otherwise: 1 Dirham is higher than 500 Fils. How is this solved in future?

One can filter for a specific currency and we have Dirham and Fils as two separate "currencies". If this becomes an important issue in the future, I'll be able to add this to the sort.

[Alt03]

Very useful

[Matfeynisya]

The current sort by Face Value

• is by descending order [ascending order since 1 Dec]

• ignores the second part of non-decimal mixed values, such as 1'6 [fixed 1 Dec - see next post]

• does not take the currency of the values into account.

For example

Sort by Face Value of https://colnect.com/en/stamps/list/coun ... /year/1964

  produces 10 s, 9 d, 7'6 s, 6 d, 5 s, 5 d, 5 s, 5 d, 2'3 s, 2 £, 2'5 s, 2'3 s, 1 £, 1'6 s

  (omitting consecutive equal values)

  after the 1 Dec 2015 changes, the same list is now in the order

                 1 £, 1'6 s, 2 £, 2'3 s, 2'5 s, 5 s, 5 d, 5 s, 5 d, 6 d, 7'6 s, 9 d, 10 s

Suggestions:

1. Sort by increasing value (I would much prefer this to descending order).

2. Take both part of non-decimal mixed values into account.

3. Assign a sort order to the currencies of each country; and then sort primarily by currency and then by face value.

For example

Sort by Face Value of https://colnect.com/en/stamps/list/coun ... /year/1964

would produce 5 d, 6 d, 9 d, 10 d; 1'6 s, 2'3 s, 2'5 s, 5 s, 7'6 s; 1 £, 2 £

Other comments

• Values such as ½ dollar would still not be sorted with 50 cents.

• Non-numeric face values such as "1st Class" and "Brev" would be sorted separately from numeric values because they have currency "none".

[tcawe]

Thanks for alerting again about this issue. I've previously not received replies to my questions.

I've now made a change to sort ascending AND to include ' character as equivalent to a numeral. Deployed, Await Restart.

This will NOT solve all sorting issues throughout and for that I'll need a clear algorithm that would convert a face value strings to something that can easily be sorted. The algorithm can be as complex as needed as I'll simply cache its results and show sorted both in the face values lists (such as stamps face values) and when sorting items.

Mind that it's not so easy to decide what should go first when we have face values represented in a myriad of different ways.

Should 7'6 be after 7.5 and before 7,8 ?

[Matfeynisya]

I have been cataloguing my stamps in my own database for some years. Here is how I have managed sorting by face value:

• I sort currencies in groups. The currency group identifier is the primary sort key.

e.g. Australian pound, Australian shilling, and Australian penny are grouped; Australian dollar and Australian cent form a separate group.

• Each currency is given a multiplying factor.

e.g. British penny=1, shilling=12, pound=240, new penny=2.4.

e.g. Indian pie=1, paisa=3, anna=12, rupee=192.

• Currencies with non-decimal subdivisions specify a second multiplying factor for the second part of the value.

e.g. Indian rupee has secondary factor 12 (for anna); Indian anna has secondary factor 1 (for pie).

• Fractions are replaced with the equivalent decimal. e.g. "½"→".5".

• The face value is separated into two parts, before and after the non-decimal separator (' in Colnect).

• The two values are multiplied by the respective factors, and added. The resulting value is the second sort key.

Using this system, each of "1 rupee, 8 anna" (entered into Colnect as "1'8 rupee"), "1.5 rupee", "1½ rupee" and "24 anna" all generate a sort value of 288.

The Australian stamp with value "1/0½ shilling" generates a sort value of 12.5.

Regarding your question "Should 7'6 be after 7.5 and before 7,8 ?":

1) My algorithm will generate the correct decimal sort value for "7'6" (which would be "7.5" if the 6 is twelfths).

2) I expect it would be very rare for decimal values such as "7.5" to be used with a currency with non-decimal subdivisions.

[tcawe]

Thank you for that info. However, implementing such a system on Colnect will be extremely time consuming. Our general approach is to list face values exactly as appearing on an item itself with the exception of separators. This explains the length of the stamp face values list.

On Colnect, we didn't create any link between the denominations of the same currency. Thus, you can see in the stamp Indian currencies list that pie / anna / paise all have different currencies.

Also consider sheets that have multiplication in the assigned face value.

Not saying that the issue cannot be resolved, it is of lower priority amongst the things we wish to implement. However, if you find the matter important and wish to help, I'd appreciate getting PHP code that would constitute a comparison function for sort or be the entire sort function, say face_value_sort(&$arFaceValues) that would get an array(face_value_id, face_value_value). It should be tested on current face values list and see that it produces coherent better results.

Thanks.

[Matfeynisya]

Yes, I agree, it would be a lot of work; and there are lots of complicating factors to take into account, some of which I have given no thought to at all.

I am very happy with the recent improvements — they make the sort by face value much more useable.

[tcawe]

If you're up to a challenge, I can send you the SQL to the face values table so you could see all the values used currently on Colnect. I've also asked one of our developers to look into it and see if he could come up with an easy solution. Thanks for raising the issue and glad that current situation is better.

[Strike26]

Dear Colnect users and stamp collectors!

I want to make a good sorting of stamps by face value.

There are some questions:

1) Please, check the list of face values

https://colnect.com/ru/stamps/face_values

and write here all the mistakes which you can find in face values.

E.g. I think that 2*FOREVER (https://colnect.com/en/stamps/list/face ... 9-2FOREVER) is not a face value and should be replaced.

One more example. There are a lot face values like 3*9.35 (https://colnect.com/en/stamps/list/face ... 12341-3935) where first factor corresponds to the number of stamps in the block. Which face value is correct?

2)I want to try the next sorting:

If I have the set of face values: 5; 5.1; 5'6; 5,3; 6+2; 0.6+3; 3*2; 3.50+11.50; 10*(39+19); 10*M

and the result of my sorting will be: 0.6+3; 3*2; 3.50+11.50; 5; 5.1; 5,3; 5'6; 6+2; 10*(39+19); 10*M

Will be this a good sorting for you? Or there should be some sorting by symbols "*" "." " +" etc?

[footy68]

E.g. I think that 2*FOREVER (https://colnect.com/en/stamps/list/face ... 9-2FOREVER) is not a face value and should be replaced.

2*FOREVER is correct. It says, we have a se-tenant of two stamps with face falue 'FOREVER'

There are a lot face values like 3*9.35 (https://colnect.com/en/stamps/list/face ... 12341-3935) where first factor corresponds to the number of stamps in the block. Which face value is correct?

3*9.35 says: 3 stamps in a block or se-tenant with face value 9.35

If I have the set of face values: 5; 5.1; 5'6; 5,3; 6+2; 0.6+3; 3*2; 3.50+11.50; 10*(39+19); 10*M

and the result of my sorting will be: 0.6+3; 3*2; 3.50+11.50; 5; 5.1; 5,3; 5'6; 6+2; 10*(39+19); 10*M

Will be this a good sorting for you? Or there should be some sorting by symbols "*" "." " +" etc?

5,3 ? the decimal sign is '.' instead of ',' in english 5,3 shouldn´t exist.

My opinion:

At first (as it is) all real numbers ,like 0.01 ....0.5 ½ ....5 ... 5.10 ....5.30 ... 10000000000.

Then all face values for non decimal face values like: 1'1 ... 1'9 .... 2'10 ... 5'6 .... 6'2 .....

Then all face values for Surtaxes like 0.05+0.02 ... 3.50+11.50 ...6+2 ... 2000000+2000000.... 0'3+1'4 ...

Then all face values for blocks,sheets,se-tenants like: 2*0.01 ... 2*1500 .... 2*FOREVER ...3*0.01 ... 3*15000 ... 3*B .... 10*0.01 ... 10*(39+19) ... 10*M...

At the end the alphabetic face values (if there is any letter inside) like 1st ... 2nd ... A... B... C.... FOREVER ... UK .... Z then kyrillic letters.

[Alt03]

Basically I agree with Guido.

I may say I like the sorting and find it very, very useful even as it is, only needs some improvements/corrections.

For example, and not pretending to be exhaustive:

- sometimes fraction precedes the correspondent decimal (e.g. 1¾ precedes 1.75) when, mostly, decimals precede fractions

- 1⁄10, 1⁄8 and 1⁄5 are after 20,000,000,000.00

- 1/3, 2⁄5, 4⁄5, 1⅗ and other fractionals are in the middle of "surtaxe" values

- the same for 1 Europe, 1st Large, 1st co and others

- the same for some multiple values, like 2*2, 4*20...

[Matfeynisya]

Here's my input on what I would like to see in the ordering of face values. I don't know if this is realistic or even implementable on Colnect.

The face value fields in Colnect have the following formats:

    base value

    base value+supplementary value

    multiple count*base value

    multiple count*(base value+supplementary value)

See https://colnect.com/en/forum/viewtopic!f=144&t=63435 for my comments regarding the syntax of face values with a multiple count.

The base value can be a currency value (numeric) or non-currency value (text).

A supplementary value is always a currency value (numeric).

A currency value can be a decimal value or non-decimal value.

A decimal value can have a decimal point and/or a fraction.

A non-decimal value has two numbers separated by '.

My preferred sort order would be:

1. Sort all currency base values before non-currency base values; sort non-currency base values in alphabetic order.

2. Sort currency values by numeric value, including sorting fractions by their numeric value; sort non-decimal second parts after values with decimal points and fractions, as in the range “1” to “2” below. (Since non-decimal values are sorted without reference to a currency, it is impossible to calculate a decimal equivalent of the whole non-decimal value.)

3. Keep all equal basic values together with supplementary values added in numeric order, as in the ranges “2” to “3” and “A” to “air” below.

4. Sort the added supplementary values in the same way as sorting the basic values, as in the range “3” to “4” below.

5. Put values with a multiple count immediately after the same values without a multiple count, as in the range “4” to “5” below.

A sample set of face values in my preferred sort order

“1” “1.40” “1.50” “1½” “1.60” “1.99” “1'1” “1'2” “1'9” “1'10” “2” “2+1” “2.50” “2.50+1” “2.50+1.50” “2.90” “2.90+1.50” “3” “3+1” “3+1.1” “3+1.5” “3+1½” “3+1.6” “3+1.9” “3+1'1” “3+1'9” “3+1'10” “4” “2*4” “4.5” “2*4.5” “4.5+2” “2*(4.5+2) ” “4.6” “5” “1st” “2nd” “2nd class” “2nd large” “3rd” “A” “A+50” “A+50000” “air” “air mail” “first-class” “Z” “Zone 1” “Б” “Ц”

Details of the sort keys to achieve this ordering

Sort Key 1

Type: text

The text of a non-currency base value; empty for a currency base value.

Sort Key 2

Type: number

The numeric value of the base value up to '; empty for non-currency base values.

Convert fractions to numbers. E.g. 10½ sorts as 10.5; 0.00½ sorts as 0.005.

Optionally add a small amount to the sort value of values with fractions so that they consistently sort after values expressed with a decimal point. E.g. Sort 10½ as 10.500001.

Sort Key 3

Type: number

The numeric value of the part of the base value after '; empty for non-currency base values, and for base values not containing '.

This must be a number to correctly sort 1'10 after 1'9 instead of between 1'1 and 1'2.

When the base value contains the non-decimal separator ', add .99999 to Sort Key 2 so that non-decimal base values sort together after all decimal base values with the same first part.

E.g. 1 1.01 … 1.99 1'1 … 1'9 1'10

This will not work correctly if the part of the base value before ' also contains a decimal, but this should not occur.

Sort Keys 4 and 5

As for keys 2 and 3 but using the supplementary value in place of the base value.

Sort Key 6

Type: number

The value of the multiple count.