| Re: declare variables [message #45934] |
Tue, 18 October 2005 16:37 |
Mark Hadfield
Messages: 783
Registered: May 1995
|
Senior Member |
|
|
qian wrote:
> Hi,
>
> I am using FORTRAN and IDL at the same time, so sometime I just assume
> some FORTRAN rules when using IDL. I just find out that the variable
> type can change within an IDL program, even you declare it to be some
> specific type. For example:
>
> data=dblarr(2)
>
> data=[2.3, 3.4]
>
> then now, 'data' is a single precision array.
>
> if I type
>
> data=[2,3]
>
> then it is an integer array.., and data(0)/data(1) = 0 !
>
> I know I should be more careful when programming, always using:
>
> data=[2.3D,4.5D]
>
> But if I read in data from other files, in which '1' actually means
> '1.0', is it possible I can re-inforce the array to be certain type (or
> let the variables to be certain type throughtout the whole program)? So
> even if I type d=1, it is still a real number?
As other respondents have pointed out, IDL is a dynamically typed
language so you can't ensure a given variable name is always associated
with data of a given type or shape. Still, there are some things you can
do to prevent unintended changes of type.
For example, consider the following
IDL> data=[2.3,3.4]
IDL> data[*] = [2,3]
IDL> print, data
2.00000 3.00000
The 1st defines data to be a 1-dimensional, 2-element, floating point
array. The 2nd assigns new values to the elements of d, but does not
change the type or shape. The 3rd confirms that the variable named data
is still floating point.
Consider another example
IDL> data = fltarr(2)
IDL> read, data
: 2 3
IDL> print, data
2.00000 3.00000
(where the colon indicates that IDL is accepting input from the
keyboard--reading from a file has the same result). Here we define data,
again as a a 1-dimensional, 2-element, floating point array, then read
in values. It doesn't matter that the numbers we read in look like
integers--the type and shape of the variable are not altered.
--
Mark Hadfield "Kei puwaha te tai nei, Hoea tahi tatou"
m.hadfield@niwa.co.nz
National Institute for Water and Atmospheric Research (NIWA)
|
|
|
|
| Re: declare variables [message #45939 is a reply to message #45934] |
Tue, 18 October 2005 14:27  |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
qian wrote:
> Hi,
>
> I am using FORTRAN and IDL at the same time, so sometime I just assume
> some FORTRAN rules when using IDL. I just find out that the variable
> type can change within an IDL program, even you declare it to be some
> specific type. For example:
You don't declare an IDL variable to have a given type. Whenever you
assign a value to an IDL variable, that variable automatically acquires
the data type of whatever it is you assigned into it.
> data=dblarr(2)
This isn't a declaration; it's an assignment statement like any other.
Since dblarr(2) is a 2-element array of doubles, that's the type of
'data'.
> data=[2.3, 3.4]
Since [2.3, 3.4] is a 2-element array of floats, that's the new type of
'data'.
Note: if you type data[0] = 2.3, then 2.3 would have been converted to
a double before storing it in data[0]. A variable's type only changes
when you assign a value to the entire variable, not when you assign to
a single element of an array.
...
> But if I read in data from other files, in which '1' actually means
> '1.0', is it possible I can re-inforce the array to be certain type (or
> let the variables to be certain type throughtout the whole program)? So
> even if I type d=1, it is still a real number?
You don't have to worry about that; it doesn't work that way. When
you're reading data from a text file using commands like:
READF, infile, myarray
Then myarray must already exist, which means it already has a specific
type, size, and array shape. The interpretatation of the text file is
controlled by the data type and size of myarray. If myarray is
currently an array of 26 integers, the above command will try to
interpret the file as containing 26 integers (and it will complain if
that interpretation doesn't work). If myarray is a 4x4 array of double
precision numbers, it will attempt to interpret the file as containing
16 floating point numbers.
|
|
|
|
| Re: declare variables [message #45940 is a reply to message #45939] |
Tue, 18 October 2005 14:07 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
qian wrote:
>
> But if I read in data from other files, in which '1' actually means
> '1.0', is it possible I can re-inforce the array to be certain type (or
> let the variables to be certain type throughtout the whole program)? So
> even if I type d=1, it is still a real number?
Hi,
You can always use the built-in FIX function on the data to recast it as
needed. See the TYPE keyword for fix.
Cheers,
Ben
|
|
|
|
| Re: declare variables [message #45941 is a reply to message #45940] |
Tue, 18 October 2005 13:35 |
Ken Mankoff
Messages: 158
Registered: February 2000
|
Senior Member |
|
|
On Tue, 18 Oct 2005, qian wrote:
> is it possible I can re-inforce the array to be certain type (or
> let the variables to be certain type throughtout the whole
> program)? So even if I type d=1, it is still a real number?
No. IDL is a very loosly typed language. You can do:
x = 3.14159
x = "sort of pi"
And there exist no mechanism to produce an error, warning, message,
or anything that the variable has changed types.
-k.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
