is variablize or variabilize a word?

Question

Is there a more appropriate word to use when ‘variablizing’ a data point (i.e to make something a variable) since the word variablize (or variabilize) does not seem to appear in the Oxford English Dictionary (yet), and it's a word people might use in certain contexts.

Answer 1

Variablize implies to make something into a variable, but only makes immediate sense to software programmers. Since it is not a dictionary word, it must be relegated to a protologism.

That being said, the closest word to your intended meaning is dynamize:

1: to make (a drug) effective (as by comminution or dilution)

2: to make dynamic: endow with force

It is the second definition -- to make dynamic -- that most readily applies here. If you make something static (i.e. a data point) into something dynamic (i.e. a variable) you dynamize it.

Of course, no one really uses dynamize in casual conversation either, even though it is in the dictionary. Don't let that stop you from trying to make variablize into a neologism, though.

Answer 2

Depending on the context (computer science, data science), I'd use the word "parameterize".

Answer 3

Two questions seem to emerge, one of which is about the English spelling of a verb derived from Latin. I shall leave the question of whether it is the right word for the particular use under discussion to people with better knowledge of the language of statistics than I have.

As to the spelling, first it seems that there is no help from academic dictionaries: neither Merriam Webster UK not Cambridge English (Brit) recognises either. But if there were such a word, or if it were to be coined among specialist terminology, there the guide is as follows.

Words (in fact, adjectives, though 'variable' is also a noun) ending in -able (such as 'liable') change to a noun by replacing '-le' with the noun ending '-ility. So:

• stable ---> stability
• irritable ---> irritability
• manageable ---> manageability and, of course:
• able ---> ability

Just one warning, though. US and British English part company over the final syllable of visualise. US, unlike the British, has visualize. Latin did not happen to have a letter 'z', though later Roman writers used this symbol to represent the Greek zeta, when translating Greek words into Latin. So the Latin spelling with an 's' came down into British English via both scholarship and Britain's Norman conquerors. The Americans went for the 'z' sound and spelled -ize and -ization.

Answer 4

In the domain of statistics, you would create a dummy variable to handle data points that should be in one class or another.

If your domain is also statistics, instead of "variablizing" data points, please consider Creating a dummy variable for those data points.

Here is a quote from an answer on our sister site, Cross Validated, which is for statisticians.

The inventor of dummy variables was George Boole in mid XIX century. On his book "An investigation of the laws of thought: on which are founded the mathematical theories of logic and probabilities" published on 1854 he proposes 0 and 1 as a mean to represent a class. You can look at his fascinating work on page 47 of his book on this link. http://www.archive.org/stream/investigationofl00boolrich#page/46/mode/2up

Here's a quote from the Wikipedia page on Dummy Variables:

Dummy variables are "proxy" variables or numeric stand-ins for qualitative facts in a regression model.

Answer 5

In the fields of applied mathematics or mathematical programming, you could use the term variabilize informally in explaining something at a whiteboard, just as you could use constantize, functionize or similar forms. For example, you might be assigning a different meaning to a symbol in an expression. Your audience will almost certainly understand your intent.

However, this is one of those areas where informality can conceal your larger meaning.

For example, if your intent is to relax a constraint in your model, you are saying something about the external world you are seeking to represent.

If your intent is to vary a parameter to study a family of behaviors, or to plot a family of curves, you are implicitly giving the parameter less freedom of movement.

If you make some quantity variable, you are asserting your ability to vary it. If you allow some quantity to vary you are possibly intending to vary it indirectly, as in an optimization problem.

In a written exposition, or in a more careful verbal exposition, these are nuances you might want to retain, and there are accepted expressions in all technical fields, which the experts know and use.

In a whiteboard presentation “amongst friends”, you can use a shortcut like variabilize because you are physically present to give context, and to assess whether your audience understood. Outside of this informal setting, it will sound clumsy and amateurish.

Answer 6

While I agree that variablize/ing just sounds wrong, there is a [usually] difference between this term and parameterize/ing in the way they are used.

To parameterize is to replace constants or literals in a piece of code (procedure, routine, function, method, ...) with parameters which are fed by arguments to that piece of code. For instance, when typing x=sin(.98), .98 is an argument. Within the sin function this argument is stored in a variable parameter. In this case, the argument to the sin function is parameterized.

To variablize is one step higher than parameterize. It is to replace a piece of code with arguments and parameters. For instance instead of saying (in some made up language):

EXEC ([define x; x = .98; define result; result = sin(x); return result;]); the entire scrap of code is stored in a variable and fed to the routine as a argument, like so: define inst LONGCHAR; inst = [define x; x = .98; define result; result = sin(x); return result;] EXEC (inst); $$ 0.017103393 Now pieces of code are treated as variables and can even be manipulated as variables.

Variablize isn't that common and would be [probably] be considered incorrect in terms of referring to basic arguments and parameters.

Answer 7

"Shochasticize"? In that you are assuming that your observation (data point) represents an example of a variable with a random distribution of some type.