talk.txt

Geomapping with Pyecharts(Echarts.js)

Abstract

The talk present the work in creating thousands of static map assets for data analysts and scientists who uses pyecharts, a python interface to Echarts.js.


Description


There exists a need for a readily available maps for python developers who are not yet familar with geo-information processing.
The need could simply be to draw a Choropleth map that illustrates the product sales in different regions of a country. However,
this need could potentially ruin a few weekends of this particular developer, because the amount of research work is huge.
Sometimes, this need is dropped because before the Choropleth map will be drawn, the deadline for the presentation is passed.

Hence the open source effort to satisfy this need started. Pyecharts user could instantly use the map assets of more than 200
countries of the world, and thousands of Chinese regions: 2883 countries/city districts, 370 provincial cities,
28 provinces/automous regions of China.

The talk covers the following topics:

1. the techniques used in creating the existing mapping assets
2. the integration of echarts.js with python, especially with jupyter notebook
3. the way to create your own map assets


Using javascript tool-chain in editing and preparing geojson for your python 

The talk presents using publicly available an alternative way of to bring in creating geomapping assets


Pyecharts is a python interface to Echarts, a javascript charting library. 

Introduction

Geomapping with pyecharts

2) What's advantage using pyecharts/echarts?

1) free of geo-spacial views, easier to get along with d3
2) interactivity comes for free
    a. zooming
    b. click

The existing static maps includes:

Inside China, pyecharts has

At provincal level: 23 provinces and 5 autonomous regions
At city leve: 370 cities that has districts
At county level: 2882 counties/city districts

1) 213 countries and regions in the world

And it has UK specialist mapping packages.

2) How does it work 

a. python interface
b. jupyter note interface
c. nteract interface
d. jupyterlab

3) how to create more maps

Techniques

1) UTF8 encoding -> take publicly avaiable geojson to echarts
2) Provincial maps: to join neighbouring provinces together
3) To create a regional China map:
    group all proinces togather, resolve their border, then join them together


4) To supply the mapping data to pyecharts, we need to create package

a. create pure npm package
b. create an encapsulation pkg package
c. then publish/install
d. use it.