Skip to content
Snippets Groups Projects
sim_functions.jl 2.14 KiB
Newer Older
  • Learn to ignore specific revisions
  • rhaag's avatar
    rhaag committed
    export data_functions, save_path, getValues
    import Pkg
    Pkg.activate(".")
    
    using DataFrames
    using DelimitedFiles
    
    
    
    # Dict of the functions used for evaluation
    # QUESTION Calculate Variance for all of these?
    data_type_functions = Dict(
        :served_percentage => RP.served_percentage,
        :driven_distance => RP.driven_distance,
        :requested_distance => RP.requested_distance,
        :mean_relative_delay => RP.mean_relative_delay,
       # :mean_occupancy => RP.mean_occupancy, #BUG In Reading TUples from a CSV File
    )
    
    
    
    """
     Converts a given index to a 2D Matrix Index with the given Matrix size
    """
    function getValue(index, xmin, xmax, xlen, ymin, ymax, ylen)
        xstep = (xmax-xmin)/xlen
        ystep = (ymax-ymin)/ylen
    
    
        x = (index%xlen) * xstep + xmin
        y = trunc(Int64, index/ylen) * ystep + ymin
        return x,y
    end
    
    Ruben Haag's avatar
    Ruben Haag committed
    function getLogValue(index, xmin, xmax, xlen::Int64, ymin, ymax, ylen::Int64)
    
    Ruben Haag's avatar
    Ruben Haag committed
        xs = exp.(LinRange(log(xmin), log(xmax), xlen))
        ys = LinRange(log(ymin), log(ymax), ylen)
        x = xs[(index-1)%xlen+1]
        y = ys[trunc(Int64, (index-1)/ylen)+1]
        return x,y
    end
    
    
    """
    
    """
    function getLogLogValue(index, xmin, xmax, xlen::Int64, ymin, ymax, ylen::Int64)
    
    
        xs = exp.(LinRange(log(xmin), log(xmax), xlen))
        ys = exp.(LinRange(log(ymin), log(ymax), ylen))
        println(length(xs));
        x = xs[(index-1)%xlen+1]
        y = ys[trunc(Int64, (index-1)/ylen)+1]
        return x,y
    end
    
    
    rhaag's avatar
    rhaag committed
    """
     Function for Running a RidePooling simulation with the normalized Frequency x
    """
    function simulate_rp(paths::Dict, N::Int64, x, y, t0::Float64, specs; served = 10*N, requested=10*N)
    
        #Make Model
        model=RP.get_model(;N_bus=N,ν=x/t0,specs...);
        RP.run!(model;requested=requested, served=served)
        data = Dict()
        data[:frequency] = x
        data[:dt_latest_dropoff] = y
        for (name, func) in data_type_functions
            data[name] = func(model)
        end
    
        # Save the calculated Data
        data = DataFrame(data)
    
        open(paths[:data]*"$index.csv", "w") do io
            writedlm(io, Iterators.flatten(([names(data)], eachrow(data))), ';')
        end
        #Save the model for possible later reference
        RP.savemodel(paths[:model]*"$index.model",model;route_matrix=false)
    end