Skip to main content

Table 1 Variables used in this study

From: Learning to cluster urban areas: two competitive approaches and an empirical validation

Type

Level

Description

Variables

Location

I/B

Geocoded position of the unit

LON, LAT

SES

I

Socioeconomic status

SES

B

 

AVG_SES, STD_SES

Aesthetic

I/B

Aesthetic perception index of the nearest geocoded image (individual) or averaged at the urban block level

BEAUTY, BORING,

  

DEPRESSING, LIVELY,

  

SAFE, WEALTHY

Political

I/B

Proportion of a political choice in the Constitutional plebiscite or in the Chilean Presidential Elections (first round)

PROP_APPROVE,

  

PROP_{BORIC,KAST,

  

PROVOSTE,SICHEL,

  

ENRIQUEZ,PARISI,

  

ARTES}

Surnames

I

Eliteness (α) and presence/absence

MAPUCHE, ELITE

B

of Mapuche or high-class surnames

AVG_{MAPUCHE,ELITE,

  

ALPHA}

Land use

I/B

Proportion/\(M^{2\ddagger}\) of land use of the urban block according to a land use typology

PROP_{A,C,D,E,G,H,I,

  

K,L,M,O,P,Q,S,T,V,W,Z}

  

M2_{A,C,D,E,G,H,I,

  

K,L,M,O,P,Q,S,T,V,W,Z}

Demographic

I

Sex, age, proportion of immigrants

SEX, AGE, PROP_IM

B

Age, proportion of immigrants/women

AVG_AGE, STD_AGE,

  

PROP_{IMM,WO}

  1. * A: farming, C: commerce, D: sports, E: education, F: forest, G: hotel, H: housing, I: industry, K: not encoded, L: storage, M: mining, O: business, P: government, Q: worship, S: health, T: transport, V: other, W: wasteland, Z: parking. † I: individuals, totaling 3,947,875 records; B: blocks, totaling 40,962 records. ‡ : we used the log of the sum of \(M^{2}\) per urban block to avoid over-representing big urban blocks.