How can we know that value is real of Integer

[avinash00002002]

I have a one problem. I can extract data from dimensions, and there is DIMRND set 0.5.

Now I have a lots of dimension text in my list. like. 110.0 , 1100.5 , 637.5, 3137.0
I had rounded up at the time of taking dimensions. now I called up in lisp as

(rtos dmval1 2 1). now I want to write above value in the drawing but as a
110.0 to 110.0
1100.5 to 1100.5
637.5 to 637.5
3137.0 to 3137

above mentioned list can be a min. 6 to max. 1250 or more.

Now my question is how can we determine 1100 is a intger and 637.5 is a real value

can anyone give me help to out.

Thanks,
Avinash

[hofcad]

I have a one problem. I can extract data from dimensions, and there is DIMRND set 0.5.

Now I have a lots of dimension text in my list. like. 110.0 , 1100.5 , 637.5, 3137.0
I had rounded up at the time of taking dimensions. now I called up in lisp as

(rtos dmval1 2 1). now I want to write above value in the drawing but as a
110.0 to 110.0
1100.5 to 1100.5
637.5 to 637.5
3137.0 to 3137

above mentioned list can be a min. 6 to max. 1250 or more.

Now my question is how can we determine 1100 is a intger and 637.5 is a real value

can anyone give me help to out.

Thanks,
Avinash

Dear Avinash,

See

(defun Int_Or_Real0(X)
(if (eq (type x) 'INT) (print "It is a Integer"))
(if (eq (type x) 'REAL) (print "It is a Real"))
(princ)
)

(Int_Or_Real0 1000); gives "Type is a Integer"
(Int_Or_Real0 1000.0); gives "Type is a Real"

(defun Int_Or_Real(X)
(if (/= (- x (fix x)) 0) (print "It is not a Integer")
(print "Type could be a Integer"))
)

(Int_Or_Real 1000); gives "Type could be a Integer"
(Int_Or_Real 1000.0); gives "Type could be a Integer"
(Int_Or_Real 1000.5); gives "It is not a Integer"

Regards HofCAD

[T.Willey]

The value you are going to get from the dimension will always be a real number. If you want to know if it can be represented as an integer, then one way is to get convert it to a string, and check to see if the numbers after the decimal are all zeros. Or you could see if the the integer value matches the real value, but I would go with the first option.

[kennet.sjoberg]

. . .convert it to a string, and check to see if the numbers after the decimal are all zeros. . .

That is not secure, often AutoCAD calculate numbers like
(zerop (- (fix 1230.0000000001 ) 1230.0000000001 )) --> nil
but truncated it may work
(zerop (- (fix 1230.000000 ) 1230.000000 )) --> T
(zerop (- (fix 1230.100254 ) 1230.100254 )) --> nil

: ) Happy Computing !

kennet

[T.Willey]

That is not secure, often AutoCAD calculate numbers like
(zerop (- (fix 1230.0000000001 ) 1230.0000000001 )) --> nil
but truncated it may work
(zerop (- (fix 1230.000000 ) 1230.000000 )) --> T
(zerop (- (fix 1230.100254 ) 1230.100254 )) --> nil

: ) Happy Computing !

kennet

You're example is correct, but it is not what I said in the quote you took. You can take a real and convert it to a string with 16 decimal places and check them all if you wanted to, but that isn't what the OP wanted either. For what the OP wants I stand by what I said.

[RobertB]

The value you are going to get from the dimension will always be a real number. If you want to know if it can be represented as an integer, then one way is to get convert it to a string, and check to see if the numbers after the decimal are all zeros. Or you could see if the the integer value matches the real value, but I would go with the first option.

I would not rely on string conversions.

(defun IsInteger  (x)
 (cond ((= (type x) 'INT))
       ((equal (- x (fix x)) 0.0 1e-15))))

Note that you can change the precision test (1e-15) but any greater precision than that value is likely overkill.

[T.Willey]

I would not rely on string conversions.

(defun IsInteger  (x)
 (cond ((= (type x) 'INT))
       ((equal (- x (fix x)) 0.0 1e-15))))

Note that you can change the precision test (1e-15) but any greater precision than that value is likely overkill.

That is a good way. Thanks Robert.

[kennet.sjoberg]

You're example is correct, but it is not what I said in the quote you took. You can take a real and convert it to a string with 16 decimal places and check them all if you wanted to, but that isn't what the OP wanted either. For what the OP wants I stand by what I said.

But I do not agree

(rtos 43.0 2 16 ) --> "42.99999999999999"
(rtos 81.0 2 16 ) --> "80.99999999999999"
(rtos 86.0 2 16 ) --> "85.99999999999999"
(rtos 91.0 2 16 ) --> "90.99999999999999"


: ) Happy Computing !

kennet

[kennet.sjoberg]

I must say again that this task is tricky

Roberts function will also sometimes return in nil

(IsInteger (atof (rtos 43.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 81.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 86.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 91.0 2 16 )) ) ; --> nil

(IsInteger (atof (rtos 15.0 2 16 )) ) ; --> T

: ) Happy Computing !

kennet

[RobertB]

I must say again that this task is tricky

Roberts function will also sometimes return in nil

(IsInteger (atof (rtos 43.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 81.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 86.0 2 16 )) ) ; --> nil
(IsInteger (atof (rtos 91.0 2 16 )) ) ; --> nil

(IsInteger (atof (rtos 15.0 2 16 )) ) ; --> T

Why, in all that's decent and fair, are you converting a real to a string and then back to a real?! You are simply forcing floating-point errors. Which is why I advised against string conversions.

_$ (IsInteger 43.0)
T
_$ (IsInteger 81.0)
T
_$ (IsInteger 86.0)
T
_$ (IsInteger 91.0)
T
_$ (IsInteger 15.0)
T       
_$ (rtos 43.0 2 16)
"42.99999999999999"
_$ (rtos 81.0 2 16)
"80.99999999999999"
_$ (rtos 86.0 2 16)
"85.99999999999999"
_$ (rtos 91.0 2 16)
"90.99999999999999"
_$ (rtos 15.0 2 16)
"15.00000000000000"