dd_cov: dd_cov
In mpadge/distdecay: Calculate Distance Decay Functions

Description Usage Arguments Note

Calculate station-to-station covariance matrix for a given city. Upper diagonal of matrix is between all trips from i and j, so dd_cov()[i, j] holds the covariance between all trips from station i to all other stations and from station j to all other stations. Analogously, the lower diagonal (dd_cov()[j, i]) holds covariances between all trips to i and j.

1 2	dd_cov(city, lower = 0, upper = 1, osm = TRUE, neutral = FALSE, k = NULL)

`city`	City for which covariance matrix is to be extracted
`lower`	Lower limit (0-1) for distance cutoff used to calculate covariances (see details)
`upper`	Upper limit (0-1) for distance cutoff used to calculate covariances (see details)
`osm`	If `FALSE`, return straight-line distances, otherwise street network distances.
`neutral`	If `TRUE`, return a covariance matrix corresponding to a neutral spatial interaction model rather than the observed trip matrix. These neutral covariance matrices can be compared with observed ones to reveal station pairs with higher-than-expected covariances.
`k`	For neutral covariance matrices, the exponential decay parameter of the spatial interaction model. If not specified, the value is determined by maximising fit between neutral and observed models.

The directory from which trip matrices are loaded can be retrieveed with dd_get_data_dir, and set with dd_set_data_dir. Covariances can be calculated between stations lying within a defined distance range using the lower and upper parameters. For example, to calculate covariances only using the nearest half of all stations, set upper = 0.5.

mpadge/distdecay documentation built on May 24, 2019, 6:08 a.m.