Evolutions of describe about questionr

[larmarange]

If I remember correctly, the orginal describe was developped by @briatte

[briatte]

I like that idea too, but I cannot remember much of the original describe function I wrote, which was basically a port from Stata.

If the idea is to make the output useful for data exploration, I would suggest adding "obs." next to vector length, and show the percentage of missing values. I would also suggest, like memisc, "translating" factor into "Nominal", ord. factor into "Ordinal", and numeric/integer into "Numeric".

> describe(d$factor_var)
[2000 obs.] Categorie socio-professionnelle
Nominal: "Employe"    NA           "Technicien" "Technicien" "Employe"  ...
7 levels: Ouvrier specialise | Ouvrier qualifie | Technicien | Profession intermediaire 
| Cadre | Employe | Autre
NAs: 347 (17.4%)

I would even go as far as to suggest, through an argument like help = TRUE:

• the most appropriate method to view stats, e.g. freq(x)
• the most appropriate method to plot, e.g. plot(table(x))

[larmarange]

You can test this commit larmarange@b681ccf

If it's OK, I will add it to a pull request.

I kept the original possibility of providing a list of variables in case of data.frame. It's working with data.frame, data_frame and data.table.

Some examples:

> describe(hdv2003$age)
[2000 obs.] 
integer: 28 23 59 34 71 ...
min: 18 - max: 97 - NAs: 0 (0%) - 78 unique values
> describe(hdv2003$age)
[2000 obs.] 
integer: 28 23 59 34 71 ...
min: 18 - max: 97 - NAs: 0 (0%) - 78 unique values
> describe(hdv2003)
[2000 obs. x 20 variables] data.frame

$id: 
integer: 1 2 3 4 5 ...
min: 1 - max: 2000 - NAs: 0 (0%) - 2000 unique values

$age: 
integer: 28 23 59 34 71 ...
min: 18 - max: 97 - NAs: 0 (0%) - 78 unique values

$sexe: 
nominal factor: "Femme" "Femme" "Homme" "Homme" "Femme" ...
2 levels: Homme | Femme
NAs: 0 (0%)

$nivetud: 
nominal factor: "Enseignement superieur y compris technique superieur" NA "Derniere annee d'etudes primaires" "Enseignement superieur y compris technique superieur" "Derniere annee d'etudes primaires" ...
8 levels: N'a jamais fait d'etudes | A arrete ses etudes, avant la derniere annee d'etudes primaires | Derniere annee d'etudes primaires | 1er cycle | 2eme cycle | Enseignement technique ou professionnel court | Enseignement technique ou professionnel long | Enseignement superieur y compris technique superieur
NAs: 112 (0.1%)

$poids: 
numeric: 2634.3982157 9738.3957759 3994.1024587 5731.6615081 4329.0940022 ...
min: 78.0783403 - max: 31092.14132 - NAs: 0 (0%) - 1877 unique values

$occup: 
nominal factor: "Exerce une profession" "Etudiant, eleve" "Exerce une profession" "Exerce une profession" "Retraite" ...
7 levels: Exerce une profession | Chomeur | Etudiant, eleve | Retraite | Retire des affaires | Au foyer | Autre inactif
NAs: 0 (0%)

$qualif: 
nominal factor: "Employe" NA "Technicien" "Technicien" "Employe" ...
7 levels: Ouvrier specialise | Ouvrier qualifie | Technicien | Profession intermediaire | Cadre | Employe | Autre
NAs: 347 (0.2%)

$freres.soeurs: 
integer: 8 2 2 1 0 ...
min: 0 - max: 22 - NAs: 0 (0%) - 19 unique values

$clso: 
nominal factor: "Oui" "Oui" "Non" "Non" "Oui" ...
3 levels: Oui | Non | Ne sait pas
NAs: 0 (0%)

$relig: 
nominal factor: "Ni croyance ni appartenance" "Ni croyance ni appartenance" "Ni croyance ni appartenance" "Appartenance sans pratique" "Pratiquant regulier" ...
6 levels: Pratiquant regulier | Pratiquant occasionnel | Appartenance sans pratique | Ni croyance ni appartenance | Rejet | NSP ou NVPR
NAs: 0 (0%)

$trav.imp: 
nominal factor: "Peu important" NA "Aussi important que le reste" "Moins important que le reste" NA ...
4 levels: Le plus important | Aussi important que le reste | Moins important que le reste | Peu important
NAs: 952 (0.5%)

$trav.satisf: 
nominal factor: "Insatisfaction" NA "Equilibre" "Satisfaction" NA ...
3 levels: Satisfaction | Insatisfaction | Equilibre
NAs: 952 (0.5%)

$hard.rock: 
nominal factor: "Non" "Non" "Non" "Non" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$lecture.bd: 
nominal factor: "Non" "Non" "Non" "Non" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$peche.chasse: 
nominal factor: "Non" "Non" "Non" "Non" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$cuisine: 
nominal factor: "Oui" "Non" "Non" "Oui" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$bricol: 
nominal factor: "Non" "Non" "Non" "Oui" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$cinema: 
nominal factor: "Non" "Oui" "Non" "Oui" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$sport: 
nominal factor: "Non" "Oui" "Oui" "Oui" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$heures.tv: 
numeric: 0 1 0 2 3 ...
min: 0 - max: 12 - NAs: 5 (0%) - 30 unique values
> describe(hdv2003, "cuisine", "heures.tv")
[2000 obs. x 2 variables] data.frame

$cuisine: 
nominal factor: "Oui" "Non" "Non" "Oui" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

$heures.tv: 
numeric: 0 1 0 2 3 ...
min: 0 - max: 12 - NAs: 5 (0%) - 30 unique values
> describe(hdv2003, "trav*")
[2000 obs. x 2 variables] data.frame

$trav.imp: 
nominal factor: "Peu important" NA "Aussi important que le reste" "Moins important que le reste" NA ...
4 levels: Le plus important | Aussi important que le reste | Moins important que le reste | Peu important
NAs: 952 (0.5%)

$trav.satisf: 
nominal factor: "Insatisfaction" NA "Equilibre" "Satisfaction" NA ...
3 levels: Satisfaction | Insatisfaction | Equilibre
NAs: 952 (0.5%)
> describe(hdv2003, "trav|lecture")
[2000 obs. x 3 variables] data.frame

$trav.imp: 
nominal factor: "Peu important" NA "Aussi important que le reste" "Moins important que le reste" NA ...
4 levels: Le plus important | Aussi important que le reste | Moins important que le reste | Peu important
NAs: 952 (0.5%)

$trav.satisf: 
nominal factor: "Insatisfaction" NA "Equilibre" "Satisfaction" NA ...
3 levels: Satisfaction | Insatisfaction | Equilibre
NAs: 952 (0.5%)

$lecture.bd: 
nominal factor: "Non" "Non" "Non" "Non" "Non" ...
2 levels: Non | Oui
NAs: 0 (0%)

> describe(femmes)
[2000 obs. x 17 variables] tbl_df tbl data.frame

$id_femme: Identifiant de l'enquêtée
integer: 391 1643 85 881 1981 ...
min: 1 - max: 2000 - NAs: 0 (0%) - 2000 unique values

$id_menage: Identifiant du ménage
integer: 381 1515 85 844 1797 ...
min: 1 - max: 1814 - NAs: 0 (0%) - 1814 unique values

$poids: Poids statistique
numeric: 1.80315 1.80315 1.80315 1.80315 1.80315 ...
min: 0.044629 - max: 4.396831 - NAs: 0 (0%) - 351 unique values

$date_entretien: Date de passation du questionnaire
Date: 2012-05-05 2012-01-23 2012-01-21 2012-01-06 2012-05-11 ...
min: 2011-12-01 - max: 2012-05-31 - NAs: 0 (0%) - 165 unique values

$date_naissance: Date de naissance
Date: 1997-03-07 1982-01-06 1979-01-01 1968-03-29 1986-05-25 ...
min: 1962-02-07 - max: 1997-03-13 - NAs: 0 (0%) - 1740 unique values

$age: Âge révolu (en années) à la date de passation du questionnaire
numeric: 15 30 33 43 25 ...
min: 14 - max: 49 - NAs: 0 (0%) - 36 unique values

$milieu: Milieu de résidence
labelled numeric: 2 2 2 2 2 ...
2 labels: [1] urbain [2] rural
min: 1 - max: 2 - NAs: 0 (0%) - 2 unique values

$region: Région de résidence
labelled numeric: 4 4 4 4 4 ...
4 labels: [1] Nord [2] Est [3] Sud [4] Ouest
min: 1 - max: 4 - NAs: 0 (0%) - 4 unique values

$educ: Niveau d'éducation
labelled numeric: 0 0 0 0 1 ...
4 labels: [0] aucun [1] primaire [2] secondaire [3] supérieur
min: 0 - max: 3 - NAs: 0 (0%) - 4 unique values

$travail: A un emploi ?
labelled numeric: 1 1 0 1 1 ...
2 labels: [0] non [1] oui
min: 0 - max: 9 - NAs: 0 (0%) - 3 unique values

$matri: Statut matrimonial
labelled numeric: 0 2 2 2 1 ...
6 labels: [0] célibataire [1] mariée [2] en concubinage [3] veuve [4] divorcée [5] séparée
min: 0 - max: 5 - NAs: 0 (0%) - 6 unique values

$religion: Religion
labelled numeric: 1 3 2 3 2 ...
5 labels: [1] musulmane [2] chrétienne [3] protestante [4] sans religion [5] autre
min: 1 - max: 5 - NAs: 4 (0%) - 6 unique values

$journal: Lit la presse ?
labelled numeric: 0 0 0 0 0 ...
2 labels: [0] non [1] oui
min: 0 - max: 1 - NAs: 0 (0%) - 2 unique values

$radio: Ecoute la radio ?
labelled numeric: 0 1 1 0 0 ...
2 labels: [0] non [1] oui
min: 0 - max: 1 - NAs: 0 (0%) - 2 unique values

$tv: Regarde la télévision ?
labelled numeric: 0 0 0 0 0 ...
2 labels: [0] non [1] oui
min: 0 - max: 1 - NAs: 0 (0%) - 2 unique values

$nb_enf_ideal: Nombre idéal d'enfants
labelled numeric: 4 4 4 4 4 ...
1 labels: [96] Ne sait pas
min: 0 - max: 99 - NAs: 0 (0%) - 18 unique values

$test: A déjà fait un test de dépistage du VIH ?
labelled numeric: 0 9 0 0 1 ...
2 labels: [0] non [1] oui
min: 0 - max: 9 - NAs: 0 (0%) - 3 unique values

[briatte]

Looks pretty good to me, very helpful output that immediately shows things that would require two or three functions to get in base R.

[larmarange]

I have prepared a new pull request with labelled functions, freq, lookfor, describe and ltabs

#54

[larmarange]

cf. Pull Request #57

[larmarange]

cf. #72