| Re: Oooh...! It's harder than I thought!! To get average of each line using only meaningful data [message #54947] |
Tue, 24 July 2007 08:52 |
Conor
Messages: 138
Registered: February 2007
|
Senior Member |
|
|
On Jul 24, 11:40 am, "edward.s.mei...@aero.org" <mei...@aero.org>
wrote:
> On Jul 24, 8:47 am, Conor <cmanc...@gmail.com> wrote:
>
>
>
>> totals = total(data_arr,1)/total(data_arr<1,1)
>
>> total(data_arr,1) returns a row vector where each element is the total
>> of a single line of data. total(data_arr<1,1) does essentially the
>> same thing, but first it imposes a ceiling on the array, so that no
>> value is greater than 1. Then, when you total across a line, what you
>> get is the total number of non-zero elements.
>
> Very clever, Conor! It just needs one small change: You have assumed
> that the total number of non-zero elements is, well, non-zero. If
> there are no non-zero elements, it will choke, so just do,
>
> totals = total(data_arr,1)/(total(data_arr<1,1) > 1)
>
> I use the */(* > 1) trick quite frequently.
>
> Ed
Good call. I always forget to check for things like that :) On a
related note, I've also assumed that the data is integer data. This
would fail for floats (or, at least for any values between 0 and 1)
|
|
|
|
| Re: Oooh...! It's harder than I thought!! To get average of each line using only meaningful data [message #54948 is a reply to message #54947] |
Tue, 24 July 2007 08:40  |
edward.s.meinel@aero.
Messages: 52
Registered: February 2005
|
Member |
|
|
On Jul 24, 8:47 am, Conor <cmanc...@gmail.com> wrote:
>
> totals = total(data_arr,1)/total(data_arr<1,1)
>
> total(data_arr,1) returns a row vector where each element is the total
> of a single line of data. total(data_arr<1,1) does essentially the
> same thing, but first it imposes a ceiling on the array, so that no
> value is greater than 1. Then, when you total across a line, what you
> get is the total number of non-zero elements.
Very clever, Conor! It just needs one small change: You have assumed
that the total number of non-zero elements is, well, non-zero. If
there are no non-zero elements, it will choke, so just do,
totals = total(data_arr,1)/(total(data_arr<1,1) > 1)
I use the */(* > 1) trick quite frequently.
Ed
|
|
|
|
| Re: Oooh...! It's harder than I thought!! To get average of each line using only meaningful data [message #54949 is a reply to message #54948] |
Tue, 24 July 2007 05:47 |
Conor
Messages: 138
Registered: February 2007
|
Senior Member |
|
|
On Jul 24, 5:47 am, weitk...@esrf.fr wrote:
> On Jul 24, 10:32 am, Ingo von Borstel <newsgro...@planetmaker.de>
> wrote:
>
>> where_notzero = WHERE(data_arr NE 0,n_notzero)
>> average = total(data_arr) / n_notzero
>
> Or simply,
>
> average = TOTAL(data_arr) / TOTAL(data_arr NE 0).
>
> Cheers,
> Timm
It seems to me that the above two examples both find the average over
the entire data array, not each row individually. To do that, I would
do something like this:
totals = total(data_arr,1)/total(data_arr<1,1)
total(data_arr,1) returns a row vector where each element is the total
of a single line of data. total(data_arr<1,1) does essentially the
same thing, but first it imposes a ceiling on the array, so that no
value is greater than 1. Then, when you total across a line, what you
get is the total number of non-zero elements. So, you are dividing
the sum of every line by the number of non-zero elements in every line
(the average). The result is a row vector, not a column vector
though. If that's a problem, you can always:
totals = reform(totals,1,numlines)
Also, I've assumed that data_arr is the same thing as in the above
examples. I would make a slight change to the reading part of the
above example though. I would change this line:
FOR i=0,n_datastrgs-1 DO data_arr[i] = DOUBLE(data_strgs[i])
To this:
data_arr = double(data_strgs)
There's no need for the for loop.
|
|
|
|
|
|
| Re: Oooh...! It's harder than I thought!! To get average of each line using only meaningful data [message #54955 is a reply to message #54954] |
Tue, 24 July 2007 01:32 |
Ingo von Borstel
Messages: 54
Registered: September 2006
|
Member |
|
|
Hello,
> test00.txt
> 0 2 4 6 7 0
> 0 0 3 3 2 0
> 1 2 3 4 5 0
> 0 0 0 1 2 0
>
> I need to get the average of each line.
> For example, I have 4 meaningful data in the first line. In this case
> I want to have 4.75 (19/4) instead of 3.167 (19/6).
>
> To do this
> 1. I need to get the number of meaningful data, n, in each line first.
> 2. summation of each line, sum.
> 3. sum/n..., and then print to output.txt
> That's all I need. I believe I can handle step 2 and 3, but step 1 is
> harder than I thought...
>
> Please give me any suggestions. Thank you !!!
Assuming you have a fixed amount of data per line, it should work
somewhere along these lines:
data_arr = dblarr(N)
OPEN, in_lun, 'test00.txt', /GET_LUN
OPENW, out_lun, 'output.txt', /GET_LUN
WHILE NOT EOF(in_lun) DO BEGIN
; read all your data, also the irrelevant, of one line and store them in
data_arr
; something like this:
data_strg = readf(in_lun,format='(a)'
data_strgs = strsplit(data_strg,' ',/EXTRACT)
n_datastrgs = N_ELEMENS(data_strgs)
FOR i=0,n_datastrgs-1 DO data_arr[i] = DOUBLE(data_strgs[i])
; now get the relevant ones
where_notzero = WHERE(data_arr NE 0,n_notzero)
average = total(data_arr) / n_notzero
; write away your data to output.txt
PRINTF, out_lun, STRING(average,FORMAT='(f10.5)')
ENDWHILE
CLOSE, out_lun
CLOSE, in_lun
Best regards,
Ingo
--
Ingo von Borstel <newsgroups@planetmaker.de>
Public Key: http://www.planetmaker.de/ingo.asc
If you need an urgent reply, replace newsgroups by vgap.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
