flowanalysis {cartograflow}

Computation of a global concentration criterion of flows values or features

Description

Computation of a global selection criterion for filtering flows values or flow features.
To be use after flowgini and before flowmap.

Usage

flowanalysis(tab, fij = NULL, critflow, critlink, result)

Arguments

tab

input flow dataset from flowgini

fij

flow value between origin and destination places

critflow

desired level of information significativity. See Details.

critlink

desired level of features density. See Details.

result

resulting filtering criterion value. See Details.

Details

-critflow = desired level of flow's information significativity (e.g. 80 -critlink = desired level of flow's features density (e.g. 20 of the flow features that represents the more significant information.

-result="density" returns the desired level of features density as a -result = "significativity" returns the level of flow significativity as a

References

Bahoken Françoise, 2016,« La cartographie d’une sélection globale de flux, entre ‘significativité’ et ‘densité’ », Netcom Online, 30-3/4 | 2016, Online since 23 March 2017, connection on 05 May 2019. URL : http://journals.openedition.org/netcom/2565 ; DOI : 10.4000/netcom.2565

Examples

library(cartograflow)
data(flowdata)

# 1/4: Computes Gini's coefficent
tabgini <- flowgini(ODpts = flows, origin = "i", destination = "j",
                     valflow = "Fij", lorenz.plot = FALSE)
### [1] Gini's coefficent = 73.16 %

# 2/4: Plot Lorenz curve
flowgini(tabgini,
        origin = "i", dest = "j", valflow = "Fij",
        lorenz.plot = TRUE
)

# 3/4: Compute critflow filtering parameter
# critflow = 0.8 #selected criterion
flowanalysis(tabgini, critflow = 0.8, result = "signif")
### [1] "threshold =  11238  ---  flows =  80 % ---  links =  22.94 %"