MathGroup Archive 2011

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Unprotect

  • To: mathgroup at
  • Subject: [mg120900] Re: Unprotect
  • From: "Oleksandr Rasputinov" <oleksandr_rasputinov at>
  • Date: Sun, 14 Aug 2011 20:19:12 -0400 (EDT)
  • Delivered-to:
  • References: <j28elu$3ev$>

On Sun, 14 Aug 2011 13:22:22 +0100, Themis Matsoukas <tmatsoukas at>  

> I want to redefine a build in command. I use the following example:
> Unprotect[Entropy];
> Entropy[x_] := 1/x
> Entropy[10.]
> When I execute these commands on a fresh kernel I get:
> SetDelayed::write: Tag Entropy in Entropy[x_] is Protected. >>
> $Failed
> Entropy::targ: Argument 10.` at position 1 should be a List, a  
> SparseArray, or a String. >>
> Entropy[10.]
> If I execute a second time I get my definition to work but I still get a  
> warning about the arguments:
> Entropy::targ: Argument 10.` at position 1 should be a List, a  
> SparseArray, or a String. >>
> 0.1
> So, two questions:
> 1. Do I have to execute twice to cause a custom definition to stick?
> 2. How do I get rid of the warning once the custom definition works?
> Running 8.01 on Lion.
> {"8.0 for Mac OS X x86 (64-bit) (February 24, 2011)", 1}
> Themis

This is a bit of an awkward situation. The reason your Unprotect did not  
work is because Entropy is an AutoLoad function and is redefined when  
first referenced. In a new session:

ClearAttributes[Entropy, ReadProtected];
?? Entropy



Entropy :=
    Hold[GeometricMean, HarmonicMean, ContraharmonicMean,
     RootMeanSquare, MeanDeviation, MedianDeviation, Skewness,
     Kurtosis, TrimmedMean, Quartiles, InterquartileRange,
     QuartileDeviation, QuartileSkewness, Commonest, Covariance,
     Correlation, Standardize, Entropy], "Statistics`Descriptive`"] /;

So the key is to reference Entropy before Unprotecting. In another new  

ClearAttributes[Entropy, ReadProtected];
?? Entropy



Entropy[Statistics`DescriptiveDump`args___] :=
  Block[{Statistics`DescriptiveDump`res =
   Statistics`DescriptiveDump`res /;
    Statistics`DescriptiveDump`res =!= $Failed]


which is the proper definition which we may now Unprotect and update:

In[6] :=
Entropy[x_] := 1/x;


During evaluation of In[6]:= Entropy::targ: Argument 10.` at position 1  
should be a List, a SparseArray, or a String. >>

Out[8] =

leaving us the message to deal with. This occurs because the order of  
DownValues matters, and your DownValue was defined after the built-in one.  
Mathematica tries to re-order DownValues so that the more specific cases  
are first, but in this case it does not seem to have succeeded, probably  
because it is not very clear which of the two DownValues should be  
considered more specific than the other. At any rate, one can re-order the  
DownValues manually:

DownValues[Entropy] = Reverse@DownValues[Entropy, Sort -> False];

Now we have:

In[10] :=

Out[10] =

which is as desired.

Note that the above technical explanation is not intended as a commentary  
on whether or not it is a good idea to re-define built-in symbols in this  
way. In general, it is better not to do so unless you have a good reason  
for it since many Mathematica functions are interdependent and redefining  
them can lead to unexpected problems.

  • Prev by Date: Re: Getting the Date and Time From a Time Server
  • Next by Date: Re: Cataloging apps for Finder files? (Mathematica notebooks especially)
  • Previous by thread: Re: Unprotect
  • Next by thread: Map legend