Estimado Damian:

Dime que tipo de datos estas manejando, series temporales, imagenes, etc?. Tienes varias opciones dependiendo de esos datos.

Saludos

Humberto

Kindly refer our papers

Climate predictability Index, GRL, v.24 no.10 pp 1239-1242 May 1997

Fractal dimension analysis of Indian climate dynamics, Chaos, Solitions and Fractals 19 (2004) 285-291

Estimado Humberto:

Estoy trabajando con datos extraídos de una encuesta de salud a pacientes que sufrieron ICTUS. La misma recoge datos como salario, nivel de escolaridad, peso, talla, y valores cualitativos como hábitos alimenticios, momento en que comienzan a trabajar y dejan el hogar, etc.

¿Puedes darme algún consejo?

Saludos, Damian

No he trabajado esos tipos de datos!!

Is your quantitative data associated with the qualitative data?  For example, are separate data blocks associated with each qualitative value or parameter or do the qualitative values change with time ?

I recently learned about a visibility graph:

http://www.pnas.org/content/105/13/4972.full

and I thought it is very interesting, as it converts fractality found in quantitative data into scale free networks. Perhaps it could be applied to qualitative data too.

Dear Stephen Warren:

Quantitative and qualitative data are separated.

Damian - in this case what is the relationship between the qualitative and quantitative data?

For example I have explored the boundaries of the Mandelbrot Set by plotting successive values of Z in the complex plain and gradually increasing the starting values of x and (i)y.   Each value is connected to the previous value by a line simply to show the trajectory, it does not imply that you can interpolate intermediate values.  You can see from the attached examples that number pairs that fall inside the Set have trajectories that approach a limiting value asymptotically as a "Strange Attractor".  Those outside quickly rush off to infinity, taking longer to do so as they approach the boundary.

Dear Mohamed El Naschie: Do you recommend me papers and books about the four points chaos game?

Damian, It is not entirely obvious to me what you mean by 'fractal pattern'. If you mean the set of transformations which generate given fractal object, your question is a hard one and requires solving an 'inverse problem'. A similar question has been asked before and you can find some answers there, if this is what you mean, that is: https://www.researchgate.net/post/Can_we_re-generate_similar_object_using_fractal_dimension_that_we_got_after_analysis

Herewith a newway of defining and seeing fractals from big data:

https://www.researchgate.net/publication/270634544_HeadTail_Breaks_for_Visualization_of_City_Structure_and_Dynamics

Article Head/tail Breaks for Visualization of City Structure and Dynamics

Damian -  I see you are looking a socio-economic data.  The first thing I would do is try and assign a numerical value to each item of qualitatitve data in each category (nivel de escolaridad, peso, talla, y valores cualitativos como hábitos alimenticios, momento en que comienzan a trabajar y dejan el hogar, etc.).  Then I would try first a scatter plot to see if there's any correlation between, say, salary and any of these categories.  If you cannot give a numerical value to the data in any category then you may have to classify them as either + (belonging to that category) or - (not belonging) and compare the groups.  The criteria for separating + and - is usuallly arbitrary and can be changed.  Testing for chaotic relationships depends absolutely on having numerical data.   There are methods of analysing the distribution of values, depending on whether the distribution is symmetrical (Poisson) or asymmetric (use Weibull).