Pushkin as a Traveller¶

@file:DependsOn("org.geotools:gt-shapefile:[30,)")
@file:DependsOn("org.geotools:gt-cql:[30,)")

import org.geotools.data.shapefile.ShapefileDataStoreFactory
import org.geotools.data.simple.SimpleFeatureCollection
import java.net.URL

import org.geotools.filter.text.cql2.CQL

val factory = ShapefileDataStoreFactory()

val worldFeatures : SimpleFeatureCollection = with("naturalearth_lowres") {
    val url = "https://raw.githubusercontent.com/JetBrains/lets-plot-kotlin/master/docs/examples/shp/${this}/${this}.shp"
    factory.createDataStore(URL(url)).featureSource.features
}
val currentZone = worldFeatures.subCollection(CQL.toFilter("continent = 'Europe' OR continent = 'Asia' or continent = 'Africa'"))

fun <T> appendStayTime(placesDf: DataFrame<T>, movesDf: DataFrame<T>): DataFrame<T> {
    fun cityNameToYears(name: String): String {
        val departureYears: List<Int> = movesDf.sortBy("year").filter { it["departure"] == name }["year"].toList() as List<Int>
        val rawArrivalYears: List<Int> = movesDf.sortBy("year").filter { it["arrival"] == name }["year"].toList() as List<Int>
        val arrivalYears: List<Int> = if (!departureYears.isEmpty() && departureYears.min() < rawArrivalYears.min()) {
            listOf(departureYears.min()) + rawArrivalYears
        } else {
            rawArrivalYears
        }
        val allYearRanges: List<Pair<Int, Int>> = arrivalYears.map { arrivalYear ->
            val departureYear = if (departureYears.isEmpty()) {
                arrivalYear
            } else {
                departureYears.firstOrNull { it >= arrivalYear } ?: arrivalYear
            }
            Pair(arrivalYear, departureYear)
        }.sortedByDescending { it.second - it.first }
        val yearRanges: MutableList<Pair<Int, Int>> = mutableListOf()
        for (yearRange in allYearRanges) {
            val outerRange = yearRanges.firstOrNull { bigYearRange: Pair<Int, Int> -> bigYearRange.first <= yearRange.first && yearRange.second <= bigYearRange.second }
            if (outerRange == null) yearRanges.add(yearRange)
        }
        return yearRanges.sortedBy { it.first }.map { yearRange ->
            if (yearRange.first == yearRange.second) {
                yearRange.first.toString()
            } else {
                "${yearRange.first}-${yearRange.second}"
            }
        }.joinToString(", ")
    }
    return placesDf.add("years") { cityNameToYears(it["name"] as String) }
}

val citySize = 3
val bigCitySize = 6
val bigCities = listOf("Moscow", "Petersburg")

val rawPlacesDf = DataFrame.readCSV("https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/pushkin/places.csv")

val rawMovesDf = DataFrame.readCSV("https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/pushkin/moves.csv")

val movesDf = rawMovesDf.join(rawPlacesDf) { "departure" match "name" }
    .rename("longitude" to "from_lon").rename("latitude" to "from_lat")
    .join(rawPlacesDf) { "arrival" match "name" }
    .rename("longitude" to "to_lon").rename("latitude" to "to_lat")
    .add("size_start") { if (it["departure"] in bigCities) { bigCitySize } else { citySize } }
    .add("size_end") { if (it["arrival"] in bigCities) { bigCitySize } else { citySize } }

val placesDf = appendStayTime(rawPlacesDf, rawMovesDf)
    .add("size") { if (it["name"] in bigCities) { bigCitySize } else { citySize } }

val moscowDf = rawPlacesDf.filter { it["name"] in bigCities }
val labelsDf1 = rawPlacesDf.filter { it["name"] in listOf("Nizhny Novgorod", "Novocherkassk", "Orenburg", "Stavropol", "Tiflis", "Vladikavkaz") }
val labelsDf2 = rawPlacesDf.filter { it["name"] in listOf("Chișinău", "Erzurum", "Kiev", "Pskov", "Simferopol", "Uralsk", "Vitebsk", "Oryol", "Ekaterinoslav") }