bicycleStatus: bicycleStatus
In fozy81/bikr: Bicycle infrastructure classification

Description Usage Arguments Format Details Value Warning Examples

bicycleStatus is used to calculate and classify an index for bicycle infrastructure based on a dataframe of bicycle indicators.

1 2	bicycleStatus(x,amsterdamIndex=TRUE,effort=TRUE,bicycleParkingWeight=1.5, routeWeight=0.8,cyclePathWeight=4,ruralWeight=2)

`x`	Dataframe input from bicycleData function
`amsterdamIndex`	Calculate status relative to data for Amsterdam included in package (default=TRUE)
`effort`	Estimate sampling effort to assess confidence in data quality (default=TRUE)
`bicycleParkingWeight`	Variable for weighting of bicycleparking input variable which is reflected in output 'cyclepath to road ratio norm weighted' (default=1.5)
`routeWeight`	Variable for weighting of routes input which is reflected in output 'bicycle route to road ratio norm' (default=0.8)
`cyclePathWeight`	Variable for weighting of cyclepath input variable which is reflected in output 'cycle route to road ratio norm weighted' (default=4)
`ruralWeight`	Variable for weighting of ruralness reflected in output 'rural weighting' (default=2)

A data frame with 13 variables - column names must match but column order not important:

name: name or unique id of area
cyclepath: length in km of cycle paths
road: total length in km of paved roads
bicycleparking: number of bicycle parking spaces
area: Area in hectares of polygon being assessed
routes: length in km of proposed Nation cycle route
proposedroutes: length in km of proposed Nation cycle route
proposedcyclepath: length in km of proposed cycle path
constructioncyclepath: Length in km of cyclepath under construction
editors: Number of openstreetmap editors of bicycle parking
lasteditdate: Date of last edit in openstreetmap of bicycle parking
proposedhighways: length in km of proposed roads
constructionshighways: length in km of construction roads

Using the example dataset in this package which uses data from Scotland and compares it against Amsterdam, its possible to see that Amsterdam provides better infrastructure. The Scottish areas are classified based on 5 quintiles on a normalised scale between 0-1 (Amsterdam = 1).

This function is designed to work from indicator values obtained from OpenStreetMap data. Specifically, it has been designed to use the bicycleData function. This function returns a dataframe from an OpenStreetMap database with the arguments required for use with bicycleStatus. Other data sources or supplementary data sources could be used as long as the dataframe is constructed correctly for this function. There are four indicator ratios of bicycle infrastructure which are combined to give an overall classification. These are returned in a dataframe: The ratio of cyclepath length to road length, ratio of National Cycle Network routes to roads length and the amount of bicycle parking per hectare. The is also a 'ruralness weighting' applied to account for less densely populated areas which arguably have quieter roads and require less cycle infrastructure. The ruralness weighting is based on the relative amount of roads to area as a proxy for population density. See bicycleData function for detailed breakdown of the OpenStreetMap data required for this function.

Once all indicators are calculated they are given weights for Cyclepath, National Cycle Route and Bicycle parking ratios respectively. This weighting is based on expert opinion and bicycle literature (both of which are deemed to subjective). Further work at linking bicycle indicators on outcomes i.e. the % of the public travelling using bicycles is planned. Ultimately, the index is base on quantitive data but the final assigned status calculated for an area is based to some degree on the weighting given to each indicator. The approach is to iterate the classification system to more closely align the overall status classification to socio-ecomonic-enviromental outcomes.

Each entry in the dataframe is also given a 'Sampling effort'. In this context, it measures the sampling effort (number of OpenStreetMap editors) and the time of last edit to represent uncertainty in the quality of the OpenStreetMap data. However, the current method is a fudge but produces figures that appear roughly feasible. Further empirical testing is required to make this robust.

Objectives and measures are calculated to project the amount of improvement required for an area to reach the equivalent status as the highest ranked area within the dataset.

dataframe containing the following:

name: name or unique id of area
cyclepath to road ratio: A ratio of cycle path to road
cyclepath to road ratio norm: A ratio of cycle path to road normalised against max value or Amsterdam region (default)
Cycle path status: Status of cycle path based on quintiles
cyclepath to road ratio norm weighted: A ratio of cycle path to road with * 4 weighting. The cycle path to road ratio is deemed to be *4 more important than the other indexes
area to bicycle parking ratio: The number of cycle parking spaces per hectare
area to bicycle parking ratio norm: The number of cycle parking areas per hectare normalised against max value or Amsterdam region (default). This will be a value between 0-1.0. A value of 1.0 is equal to the max. If using amsterdamIndex=TRUE parameter (default), the value is capped at 1.0 even if higher values are produced than the value found in Amsterdam region. The value is capped because the Amsterdam region is being used as the benchmark.
area to bicycle parking norm weighted: A ratio of cycle parking to area. The cycle parking to road ratio is deemed to be 1.5 more important than the other indexes as default
Bicycle parking status: Status of bicycle parking based on quintiles of the normalised score. The five categories are High, Good, Moderate, Poor or Bad. The boundaries are less than or equal to 0.2 = Bad, 0.4 = Poor, 0.6 = Moderate, 0.8 = Good, 1.0 = High. For instance, if a location has 60% the parking of Amsterdam it will be categorised as 'Moderate'.
rural weighting: Rural weighting is a ratio of the length of road divided by area. This gives an idea of the road density within a given area and broadly how rural an area may be. It is thought rural areas will on balance require less bicycle infrastrucutre because quite country roads, for example single lane roads on islands and roads in remote areas, don't require off road cycle paths or cycle lanes to the same extent as busy urban roads.
bicycle route to road ratio: A ratio of National Cycle Network route to road length
bicycle route to road ratio norm: A normalised ratio of cycle route to road. The normalised value will be between 0-1.0. A value of 1.0 is equal to the max. If using amsterdamIndex=TRUE parameter (default), the value is capped at 1.0 even if values higher than Amsterdam region are discovered. The value is capped because the Amsterdam region is being used as the benchmark. Currently having more national cycle network than Amsterdam is seen as superfluous if Amsterdam is accepted as the Gold standard in bicycle infrastructure provision.
National cycle network status: Status of National Cycle Network based on quintiles of the normalised score. The five categories are High, Good, Moderate, Poor or Bad. The boundaries are less than or equal to 0.2 = Bad, 0.4 = Poor, 0.6 = Moderate, 0.8 = Good, 1.0 = High. For instance, if a location has 85% the length of Amsterdam it will categorised as 'High'.
cycle route to road ratio norm weighted: A ratio of cycle route to road ratio with * 2 weighting. The cycle path to road ratio is deemed to be *2 more important than the other indexes
Indicator total: The Indicator total is the sum of the normalised ratios
Total normalised: The Total normalised is the sum of the weighted normalised ratios which is also normalised against the max or Amsterdam Indicator total
Status: Status is based on quintiles of the Total normalised value. The five categories are High, Good, Moderate, Poor or Bad. The boundaries are less than or equal to 0.2 = Bad, 0.4 = Poor, 0.6 = Moderate, 0.8 = Good, 1.0 = High. For instance, if a location has 15% the Total normalised value of Amsterdam it will be categorised as 'Bad'.
Sampling Effort: The Sampling Effort is a percentage estimate of the sampling effort and therefore the quality of the underlying map data. It is based on three features of the bicycle parking data openstreetmap extracted: The average number of versions of each node, the total number of editors and the timestamp of the most recent edit.
Rank: A rank 1,2,3...N is attributed to each area. '1' represents the area with the highest 'Total normalised' value