Re: Some IDL_Number documentation quirks [message #93967 is a reply to message #93966] |
Wed, 07 December 2016 16:36  |
penteado
Messages: 866 Registered: February 2018
|
Senior Member Administrator |
|
|
There were also some problems in the documentation for BitGet and BitSet, which Wayne and me talked about in an earlier thread:
https://groups.google.com/d/msg/comp.lang.idl-pvwave/bM6iI_7 Y1mE/f2S0WkJ7AwAJ
On Wednesday, December 7, 2016 at 2:17:16 PM UTC-8, Chris Torrence wrote:
> On Friday, December 2, 2016 at 12:14:52 PM UTC-7, wlandsman wrote:
>> A few quirks in the IDL documentation for IDL_Number
>>
>> http://www.harrisgeospatial.com/docs/IDL_Number.html
>>
>> 1. The example for IDL_Number::Median method reads
>> *********
>> Print the median value of a number:
>>
>> num = -24601
>>
>> PRINT, num.Median( )
>>
>> IDL prints:
>>
>> 24601
>> **********
>> Huh? The median of a negative scalar is its absolute value? I then realized that this example is probably an unfinished copy/paste from the IDL_Number:Signum method. But
>>
>> 2. The example for IDL Number::Signum is incorrect
>>
>> *********************************
>> Print the signum value of a number:
>>
>> num = -24601
>>
>> PRINT, num.Signum( )
>>
>> IDL prints:
>>
>> 24601
>> ***********************************
>> This is closer but still incorrect. The value of num.signum() should be -1 (the sign of the number).
>>
>> 3. The documentation of IDL_Number::Total is incomplete.
>>
>> **************
>> Arguments
>>
>> None.
>> Keywords
>>
>> None.
>> ***********
>>
>> But in fact, although not documented, the dimension argument and five keywords available to the TOTAL() function are also available to IDL_Number::Total. For example
>>
>> IDL> a = indgen(3,3)
>> IDL> a
>> 0 1 2
>> 3 4 5
>> 6 7 8
>> IDL> a.total(1,/integer)
>> 3 12 21
>
> Whoops, my bad. Thanks Wayne for catching all of these. They'll be fixed in IDL Next (not 8.6). You can also use the Dimension argument to the IDL_Number::Product method.
> Cheers,
> Chris
> Harris Geospatial Solutions
|
|
|