{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualizing Spatial Information\n",
"\n",
"Function `geom_livemap()` enables a researcher to visualize geospatial information on interactive map.\n",
"\n",
"When building interactive geospatial visualizations with *Lets-Plot* the visualisation workflow remains the \n",
"same as it is when building a regular `ggplot2` plot.\n",
"\n",
"However, `geom_livemap()` creates an interactive base-map super-layer and certain limitations do apply \n",
"comparing to a regular `ggplot2` geom-layer:\n",
"\n",
"* `geom_livemap()` must be added as a 1-st layer in plot;\n",
"* Maximum one `geom_livemap()` layer is alloed per plot;\n",
"* Not any type of *geometry* can be combined with interactive map layer in one plot;\n",
"* Internet connection to *map tiles provider* is required.\n",
"\n",
"The following `ggplot2` geometry can be used with interactive maps:\n",
"\n",
"* `geom_point`\n",
"* `geom_rect`\n",
"* `geom_path`\n",
"* `geom_polygon`\n",
"* `geom_segment`\n",
"* `geom_text`\n",
"* `geom_tile`\n",
"* `geom_vline`, `geon_hline`\n",
"* `geom_bin2d`\n",
"* `geom_contour`, `geom_contourf`\n",
"* `geom_density2d`, `geom_density2df`\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:06.706240Z",
"iopub.status.busy": "2024-11-01T20:16:06.706130Z",
"iopub.status.idle": "2024-11-01T20:16:07.030320Z",
"shell.execute_reply": "2024-11-01T20:16:07.030014Z"
}
},
"outputs": [],
"source": [
"from lets_plot import *"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.031766Z",
"iopub.status.busy": "2024-11-01T20:16:07.031658Z",
"iopub.status.idle": "2024-11-01T20:16:07.033644Z",
"shell.execute_reply": "2024-11-01T20:16:07.033477Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" \n",
" "
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"LetsPlot.setup_html()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### The First Map"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.047535Z",
"iopub.status.busy": "2024-11-01T20:16:07.047450Z",
"iopub.status.idle": "2024-11-01T20:16:07.077820Z",
"shell.execute_reply": "2024-11-01T20:16:07.077528Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"p = ggplot() + ggsize(700, 400)\n",
"\n",
"# Add an empty base-map layer.\n",
"p + geom_livemap()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Initial location and zoom level\n",
"\n",
"The default initial location and zoom level shown on the figure above \n",
"can be adjusted by using the `location` and `zoom` parameters."
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.078938Z",
"iopub.status.busy": "2024-11-01T20:16:07.078864Z",
"iopub.status.idle": "2024-11-01T20:16:07.081618Z",
"shell.execute_reply": "2024-11-01T20:16:07.081367Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# In Lets-Plot coordinates are always encoded as: `longitude` - first and `latitude` - second\n",
"p + geom_livemap(location=[29.737809, 60.003151], zoom=11)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Adding 'geometries'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_hline(), geom_vline()\n",
"\n",
"Map's initial `location` and `zoom` parameters are not used very often because map \n",
"automatically chooses its initial viewport so that all the added *geometries* are visible.\n",
"\n",
"But the `horizontal` and `vertical` lines are exception because they do not have determined bounds. In this case its often necessary to set the desired location and zoom value manually."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.082501Z",
"iopub.status.busy": "2024-11-01T20:16:07.082426Z",
"iopub.status.idle": "2024-11-01T20:16:07.085707Z",
"shell.execute_reply": "2024-11-01T20:16:07.085531Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lons = [-73 + x for x in range(4)]\n",
"lats = [41.5 + 0.5 * y for y in range(4)]\n",
"\n",
"p + geom_livemap(location=[-71.94375, 42.572511], zoom=7)\\\n",
" + geom_hline(aes(yintercept=lats), color='red', linetype=2)\\\n",
" + geom_vline(aes(xintercept=lons), color='dark_green', linetype=4)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_path()"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.086525Z",
"iopub.status.busy": "2024-11-01T20:16:07.086456Z",
"iopub.status.idle": "2024-11-01T20:16:07.089798Z",
"shell.execute_reply": "2024-11-01T20:16:07.089632Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"path = {\n",
" 'lon': [11.620273, -71.075466, -121.898220, -96.761501], \n",
" 'lat': [48.169211, 42.383651, 37.362240, 32.797371]\n",
"}\n",
"p + geom_livemap()\\\n",
" + geom_path(data=path, mapping=aes(x='lon', y='lat'), size=1, color='magenta')\n"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.090527Z",
"iopub.status.busy": "2024-11-01T20:16:07.090453Z",
"iopub.status.idle": "2024-11-01T20:16:07.093006Z",
"shell.execute_reply": "2024-11-01T20:16:07.092844Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"p + geom_livemap()\\\n",
" + geom_path(data=path, mapping=aes(x='lon', y='lat'), size=1, color='magenta')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_point()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.093745Z",
"iopub.status.busy": "2024-11-01T20:16:07.093670Z",
"iopub.status.idle": "2024-11-01T20:16:07.096587Z",
"shell.execute_reply": "2024-11-01T20:16:07.096416Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"p + geom_livemap()\\\n",
" + geom_path(data=path, mapping=aes(x='lon', y='lat'), size=1, color='magenta')\\\n",
" + geom_point(data=path, mapping=aes(x='lon', y='lat'), shape=21, size=7, fill='yellow')\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_segment()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.097336Z",
"iopub.status.busy": "2024-11-01T20:16:07.097266Z",
"iopub.status.idle": "2024-11-01T20:16:07.100845Z",
"shell.execute_reply": "2024-11-01T20:16:07.100637Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data = dict(\n",
" x0 = [-97.8, 22.6], y0 = [41.2, -1.3],\n",
" x1 = [ 19.3, -64.5], y1 = [18.7, -15.9],\n",
" Direction = ['east', 'west']\n",
")\n",
"p + geom_livemap()\\\n",
" + geom_segment(data=data, mapping=aes(x='x0', y='y0', xend='x1', yend='y1', color='Direction'), size=1, linetype=2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_text()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.101523Z",
"iopub.status.busy": "2024-11-01T20:16:07.101453Z",
"iopub.status.idle": "2024-11-01T20:16:07.104762Z",
"shell.execute_reply": "2024-11-01T20:16:07.104594Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data = {\n",
" 'label': [\"== 0 ==>\", \"== 60 ==>\", \"== 120 ==>\", \"== 180 ==>\", \"== -60 ==>\", \"== -120 ==>\"],\n",
" 'angle': [0, 60, 120, 180, -60, -120],\n",
"}\n",
"\n",
"p + geom_livemap() + geom_point(x=0, y=0, size=140, color='white', alpha=0.6)\\\n",
" + geom_text(aes(label = 'label', angle='angle'), data=data,\n",
" x=0, y=0, \n",
" hjust = \"left\", vjust = \"center\",\n",
" size=10, family='monospace', color='red')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_polygon()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.105404Z",
"iopub.status.busy": "2024-11-01T20:16:07.105336Z",
"iopub.status.idle": "2024-11-01T20:16:07.110124Z",
"shell.execute_reply": "2024-11-01T20:16:07.109849Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"UK = {\n",
" 'lon': [-2.598046, -2.378320, -2.466210, -3.169335, -1.938867, \n",
" -0.576562, -0.312890, 0.873632, 0.082617, -2.598046], \n",
" 'lat': [51.030349, 51.797754, 53.945750, 54.561879, 55.193929, \n",
" 53.816229, 52.924809, 52.525588, 51.113188, 51.030349]\n",
"}\n",
"Germany = {\n",
" 'lon': [7.685156, 9.926367, 13.661718, 14.101171, 11.464453, \n",
" 12.870703, 8.564062, 8.651953, 6.806250, 6.938085, 7.685156], \n",
" 'lat': [53.294124, 54.049078, 53.608160, 51.305902, 50.221916, \n",
" 48.679365, 48.007575, 49.485266, 50.024691, 51.552493, 53.294124]\n",
"}\n",
"France = {\n",
" 'lon': [-2.246484, 2.367773, 7.245703, 5.268164, 6.586523, \n",
" 2.895117, 2.279882, -0.532617, -0.356835, -2.246484], \n",
" 'lat': [48.095702, 50.586036, 48.795295, 46.365136, 44.169607, \n",
" 43.663114, 43.088157, 43.631315, 46.516550, 48.095702]\n",
"}\n",
"polygons = dict(\n",
" lon = UK['lon'] + Germany['lon'] + France['lon'],\n",
" lat = UK['lat'] + Germany['lat'] + France['lat'],\n",
" country = ['UK' for _ in UK['lon']] + ['Germany' for _ in Germany['lon']] + ['France' for _ in France['lon']]\n",
")\n",
"\n",
"p + geom_livemap()\\\n",
" + geom_polygon(aes('lon', 'lat'),\n",
" data=polygons, \n",
" fill='green', \n",
" alpha=0.3)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_rect()"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.110790Z",
"iopub.status.busy": "2024-11-01T20:16:07.110714Z",
"iopub.status.idle": "2024-11-01T20:16:07.114202Z",
"shell.execute_reply": "2024-11-01T20:16:07.114031Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"bounds = dict(\n",
" lon0 = [min(UK['lon']), min(Germany['lon']), min(France['lon'])],\n",
" lon1 = [max(UK['lon']), max(Germany['lon']), max(France['lon'])],\n",
" lat0 = [min(UK['lat']), min(Germany['lat']), min(France['lat'])],\n",
" lat1 = [max(UK['lat']), max(Germany['lat']), max(France['lat'])],\n",
" country = ['UK', 'Germany', 'France']\n",
")\n",
"\n",
"p + geom_livemap()\\\n",
" + geom_rect(aes(xmin='lon0', ymin='lat0', xmax='lon1', ymax='lat1', fill='country'),\n",
" data=bounds, \n",
" alpha = 0.3) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Data on Map\n",
"\n",
"In the examples above we've seen how values in input data series can be mapped to aesthetic values on map.\n",
"For the instance, values in the `country` series were mapped to the `fill` color in \"geom_rect()\" example.\n",
"\n",
"Lets plot another data - the countries population:"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.114839Z",
"iopub.status.busy": "2024-11-01T20:16:07.114769Z",
"iopub.status.idle": "2024-11-01T20:16:07.116075Z",
"shell.execute_reply": "2024-11-01T20:16:07.115908Z"
}
},
"outputs": [],
"source": [
"pop_UK = 66.65e6\n",
"pop_Germany = 83.02e6\n",
"pop_France = 66.99e6"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Proportional symbols map"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.116699Z",
"iopub.status.busy": "2024-11-01T20:16:07.116629Z",
"iopub.status.idle": "2024-11-01T20:16:07.120110Z",
"shell.execute_reply": "2024-11-01T20:16:07.119944Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Create data frame containing countries name, centroid and population.\n",
"centroids = dict(\n",
" lon = [(max(xs) - min(xs)) / 2 + min(xs) for xs in [UK['lon'], Germany['lon'], France['lon']]],\n",
" lat = [(max(ys) - min(ys)) / 2 + min(ys)for ys in [UK['lat'], Germany['lat'], France['lat']]],\n",
" country = ['UK', 'Germany', 'France'],\n",
" population = [pop_UK, pop_Germany, pop_France] \n",
")\n",
"\n",
"p + geom_livemap() + scale_size(range=[10, 50], guide='none')\\\n",
" + geom_point(aes('lon', 'lat', color='country', size='population'),\n",
" data=centroids, \n",
" alpha = 0.8) \n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Choropleth map"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.120775Z",
"iopub.status.busy": "2024-11-01T20:16:07.120700Z",
"iopub.status.idle": "2024-11-01T20:16:07.123934Z",
"shell.execute_reply": "2024-11-01T20:16:07.123767Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Create data frame containing countries name, boundary polygon and population.\n",
"polygons_and_pop = polygons.copy()\n",
"polygons_and_pop['population'] = [pop_UK for _ in UK['lon']]\\\n",
" + [pop_Germany for _ in Germany['lon']]\\\n",
" + [pop_France for _ in France['lon']]\n",
"\n",
"# Note that in this case it's necessary to define `group` aesthetic.\n",
"# Otherwise `geom_polygon` will not be able to split the data into tree \"data points\".\n",
"p + geom_livemap()\\\n",
" + geom_polygon(aes('lon', 'lat', fill='population', group='country'), \n",
" data=polygons_and_pop, \n",
" alpha=0.3)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Choropleth map (geopandas and `map_join`)\n",
"\n",
"The dataframe format used in the previouse `Choropleth` example is not the most \n",
"convenient way to represent data with geospatial boundaries because each\n",
"real data-point is duplicated for each vertex in the corresponding polygon.\n",
"\n",
"More convenient would be to store the data-points and their geospatial boundaries in \n",
"a separate data structures.\n",
"\n",
"In this example let's use `geopandas.GeoDataFrame` to store the boundaries."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.124554Z",
"iopub.status.busy": "2024-11-01T20:16:07.124484Z",
"iopub.status.idle": "2024-11-01T20:16:07.125851Z",
"shell.execute_reply": "2024-11-01T20:16:07.125684Z"
}
},
"outputs": [],
"source": [
"# Data-points.\n",
"population = dict(\n",
" country = [\"UK\", \"Germany\", \"France\"],\n",
" population = [pop_UK, pop_Germany, pop_France]\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.126448Z",
"iopub.status.busy": "2024-11-01T20:16:07.126379Z",
"iopub.status.idle": "2024-11-01T20:16:07.131276Z",
"shell.execute_reply": "2024-11-01T20:16:07.131109Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from geopandas import GeoDataFrame, points_from_xy\n",
"from shapely.geometry import Polygon, LinearRing\n",
"\n",
"# GeoDataFrame with boundaries.\n",
"polygons_gdf = GeoDataFrame(\n",
" data=dict(\n",
" name = ['UK', 'Germany','France']\n",
" ),\n",
" geometry=[\n",
" Polygon(tuple(zip(UK['lon'], UK['lat']))),\n",
" Polygon(tuple(zip(Germany['lon'], Germany['lat']))),\n",
" Polygon(tuple(zip(France['lon'], France['lat'])))\n",
" ]\n",
")\n",
"\n",
"# Use `map` parameter to pass `boundaries` and \n",
"# `map_join` parameter to tell `geom_polygon` how to merge \"data\" and \"map\".\n",
"p + geom_livemap()\\\n",
" + geom_polygon(aes(fill='population'), \n",
" data=population, \n",
" map=polygons_gdf, \n",
" map_join=[['country'],['name']],\n",
" alpha=0.3)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Geometries with built-in statistical transformation\n",
"\n",
"The way these geometries are added to an interactive map is the same as it is done on regular plots."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_bin2d()"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.131944Z",
"iopub.status.busy": "2024-11-01T20:16:07.131873Z",
"iopub.status.idle": "2024-11-01T20:16:07.135350Z",
"shell.execute_reply": "2024-11-01T20:16:07.135184Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import numpy as np\n",
"np.random.seed(32)\n",
"x, y = np.random.multivariate_normal(mean=[0,0], cov=[[1,0],[0,1]], size=50).T\n",
"p + geom_livemap() + scale_fill_gradient('white', 'darkgreen')\\\n",
" + geom_bin2d(aes(x * 5, y * 5, fill='..density..'), bins=[7, 7], alpha=0.5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### geom_density2df()"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"execution": {
"iopub.execute_input": "2024-11-01T20:16:07.136003Z",
"iopub.status.busy": "2024-11-01T20:16:07.135933Z",
"iopub.status.idle": "2024-11-01T20:16:07.224080Z",
"shell.execute_reply": "2024-11-01T20:16:07.223874Z"
}
},
"outputs": [
{
"data": {
"text/html": [
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"p + geom_livemap() + scale_fill_gradient('white', 'darkgreen')\\\n",
" + geom_density2df(aes(x * 5, y * 5, fill='..level..'), alpha=0.5)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.13"
}
},
"nbformat": 4,
"nbformat_minor": 4
}