Formula Needed...

[Guest]

I have a row of numbers from 1 to 26 in cells A1 to Z1 representing numbers
'chosen' at random by 500 unique people.

I have another row of numbers representing the 'frequency' that the
aforementioned number was chosen in cells A2 to Z2.

Example:

1 2 3 4 5 6 7 8 9 10 11 ...
10 4 16 9 23 20 61 13 18 43 17 ...

Is there a formula that I could use to calculate the percentage (0%-100%)
that each number was chosen?

Thanx.

 

[mbarron]

Assuming A2 to Z2 are the frequencies:
in cell A3
=A2/SUM($A$2:$Z$2)

Copy the formula to the rest of the columns.

 

[Ron Rosenfeld]

I have a row of numbers from 1 to 26 in cells A1 to Z1 representing numbers
'chosen' at random by 500 unique people.

I have another row of numbers representing the 'frequency' that the
aforementioned number was chosen in cells A2 to Z2.

Example:

1 2 3 4 5 6 7 8 9 10 11 ...
10 4 16 9 23 20 61 13 18 43 17 ...

Is there a formula that I could use to calculate the percentage (0%-100%)
that each number was chosen?

Thanx.

Shouldn't that just be

=Frequency_of_num_choices/500


--ron

 

[Guest]

Sum all your frequencies. If they exceed 500 you have a data integrity issue.

 

[Guest]

I thought so too, but when I put that formula in, I end up with values over
100%.

Thanx.

 

[Guest]

In cell A3 you could put the formula:

=A2/sum($A$2:$Z$2)

Then just copy A3 to all cells from B3 to Z3. The $ above will prevent
Excel from automatically incrementing those references as the formula is
copied over. And by using the SUM function rather than 500, you can get an
accurate percentage regardless of how many total results you have.

 

[Guest]

wouldn't it just be the frequency divided by 500?

 

[Guest]

Unfortunately, when I use that formula, all I get in the A3 to Z3 cells are
numbers like [ 0.113065327 0.045226131 0.118090452 0.170854271
0.108040201 0.005025126 0.06281407 0.047738693 0.211055276 0.118090452
....].

Assuming, based on those numbers that I need to format the cells with a
PERCENT format, the numbers all add up wrong...I end up with numbers higher
than 100%.

Any other ideas?

Thanx.

 

[Ron Rosenfeld]

What you are posting is inconsistent.

If each of your 500 people make one choice, then there are 500 choices made.
That has to be the denominator. Unless your 500 people are making more than
one choice apiece.

Also, in your example, the first entry (under the 1) was a 10; the second
entry, under the 2, was a 4.

=10/500 is 0.02; not 0.113065327
=4/500 is 0.008; not 0.045226131

So there is clearly something inconsistent between what you've posted and what
you are getting for results.

If your frequency is in A2:Z2, then in A3 try the formula:

=A2/SUM$A$2:$Z$2) and copy/drag across

Unfortunately, when I use that formula, all I get in the A3 to Z3 cells are
numbers like [ 0.113065327 0.045226131 0.118090452 0.170854271
0.108040201 0.005025126 0.06281407 0.047738693 0.211055276 0.118090452
...].

Assuming, based on those numbers that I need to format the cells with a
PERCENT format, the numbers all add up wrong...I end up with numbers higher
than 100%.

Any other ideas?

Thanx.

In cell A3 you could put the formula:

=A2/sum($A$2:$Z$2)

Then just copy A3 to all cells from B3 to Z3. The $ above will prevent
Excel from automatically incrementing those references as the formula is
copied over. And by using the SUM function rather than 500, you can get an
accurate percentage regardless of how many total results you have.

--ron

 

[Guest]

I guess I need to explain it better...never really was that good at math.

8^)>

Looking at all the available functions within Excel, the closest one to what
I'm looking for would have to be the RANK function. However, even that one is
not quite there. What I need is a function that, based on the numbers chosen
and the freqency of each number chosen, I'd like to display the results based
on a scale between 1% and 100% instead of based on the percent of how many
choices.

Does that make sense?

Thanx.

What you are posting is inconsistent.

If each of your 500 people make one choice, then there are 500 choices made.
That has to be the denominator. Unless your 500 people are making more than
one choice apiece.

Also, in your example, the first entry (under the 1) was a 10; the second
entry, under the 2, was a 4.

=10/500 is 0.02; not 0.113065327
=4/500 is 0.008; not 0.045226131

So there is clearly something inconsistent between what you've posted and what
you are getting for results.

If your frequency is in A2:Z2, then in A3 try the formula:

=A2/SUM$A$2:$Z$2) and copy/drag across

Unfortunately, when I use that formula, all I get in the A3 to Z3 cells are
numbers like [ 0.113065327 0.045226131 0.118090452 0.170854271
0.108040201 0.005025126 0.06281407 0.047738693 0.211055276 0.118090452
...].

Assuming, based on those numbers that I need to format the cells with a
PERCENT format, the numbers all add up wrong...I end up with numbers higher
than 100%.

Any other ideas?

Thanx.

In cell A3 you could put the formula:

=A2/sum($A$2:$Z$2)

Then just copy A3 to all cells from B3 to Z3. The $ above will prevent
Excel from automatically incrementing those references as the formula is
copied over. And by using the SUM function rather than 500, you can get an
accurate percentage regardless of how many total results you have.

:

I have a row of numbers from 1 to 26 in cells A1 to Z1 representing numbers
'chosen' at random by 500 unique people.

I have another row of numbers representing the 'frequency' that the
aforementioned number was chosen in cells A2 to Z2.

Example:

1 2 3 4 5 6 7 8 9 10 11 ...
10 4 16 9 23 20 61 13 18 43 17 ...

Is there a formula that I could use to calculate the percentage (0%-100%)
that each number was chosen?

Thanx.

--ron

 

[Ron Rosenfeld]

I guess I need to explain it better...never really was that good at math.

8^)>

Looking at all the available functions within Excel, the closest one to what
I'm looking for would have to be the RANK function. However, even that one is
not quite there. What I need is a function that, based on the numbers chosen
and the freqency of each number chosen, I'd like to display the results based
on a scale between 1% and 100% instead of based on the percent of how many
choices.

Does that make sense?

No. At least not in conjunction with your previous posts.

If you have 500 people each making one choice amongst 26 options; and 50 of
them choose option 2; then 50/100 = 10% will have chosen option 2.

If that's not the kind of result you are looking for, then I'd need to see a
better example of what the data looks like, and what the results should look
like, than I've seen so far.


--ron

 

[Guest]

Can I zip and email you the spreadsheet in question?

 

[Ron Rosenfeld]

OK, but since I'm paranoid, what you'll need to do is reverse the string, make
the obvious substitution, and then throw out the garbage.

mocegabrag.enilnodlefnesorTAzyx

Can I zip and email you the spreadsheet in question?

--ron

 

[Ron Rosenfeld]

No. At least not in conjunction with your previous posts.

If you have 500 people each making one choice amongst 26 options; and 50 of
them choose option 2; then 50/100 = 10% will have chosen option 2.

If that's not the kind of result you are looking for, then I'd need to see a
better example of what the data looks like, and what the results should look
like, than I've seen so far.


--ron

That should be 50/500 = 10%

--ron

 

[Guest]

I have a related question maybe someone can assist with and will be of some
use here:
My data is formatted so that column 1 as the 'bin value' and column 2 as the
'frequency ' that bin value occurs in the set. e.g. if the individual data
were
5 5 5 6 6 6 6 7 7 8 8 8 8 8 etc..
my data format would be
5 3
6 4
7 2
8 5

basically the frequencies of my data have already been determined.
I now want to calculate things (mean, sd, skew,kurtosis) on my data set,but
only have this 2 column format to use. Excell needs all the data listed out.
Isthere some way to quickly determine these descriptive stats? I ask because
my data sets consist of 200-300 bin values, with frequencis between 10 and
300; and total frequencies (total number of data points in a data set ) is
about 12,000

douglas

 

[bplumhoff]

Hello Douglas,

I suggest to enter as UDF:

Function one_dim_array2(rI As Range) As Variant
'Converts an argument array of frequencies and values into
'a one-dimensional array.
'Example:
'3 25
'1 10
'4 12
'6 7
'will become 25 25 25 10 12 12 12 12 7 7 7 7 7 7.

Dim vR As Variant
Dim r As Range
Dim b As Boolean
Dim lF As Long, i As Long, j As Long

j = 0
For i = 1 To rI.Rows.Count
j = j + rI.Cells(i, 1).Value
Next i

ReDim vR(1 To j)

i = 1
b = True
For Each r In rI
If b Then
lF = r.Value
Else
Do While lF > 0
vR(i) = r.Value
i = i + 1
lF = lF - 1
Loop
End If
b = Not b
Next r

one_dim_array2 = vR

End Function


Then enter into any cell =AVERAGE(one_dim_array2($A$1:$B$4)) or other
desired functions like MEDIAN, MODE, SKEW, KURT...

Please notice that this UDF assumes frequencies first, then values
(your example is the other way round).

HTH,
Bernd