I'm struggling to find an elegant and reliable way of merging two arrays in vanilla JS or d3.js, where
- the "rows" of the series do not necessarily match;
- objects may share some but not all attributes;
- "missing" attributes are filled in as null.
For example, two arrays of the form
A =
[
{
year : 2001,
gdp : 1.1,
population : 1100
},
{
year : 2002,
gdp : 1.2,
population : 1200
},
]
B =
[
{
year : 2000,
gdp : 1.0,
rainfall : 100
},
{
year : 2001,
gdp : 1.1,
rainfall : 110
}
]
would ideally combine to give
C =
[
{
year : 2000,
gdp : 1.0,
population : null,
rainfall : 10
},
{
year : 2001,
gdp : 1.1,
population : 1100,
rainfall : 110
},
{
year : 2002,
gdp : 1.2,
population : 1200,
rainfall : null
},
]
I can usually figure this kind of thing out but this one has got me really stuck!
CodePudding user response:
You could build an array of all items, get the keys and build an empty object as pattern for an empty object. Then group by year
and get the values. Apply sorting if necessary.
const
a = [{ year: 2001, gdp: 1.1, population: 1100 }, { year: 2002, gdp: 1.2, population: 1200 }],
b = [{ year: 2000, gdp: 1.0, rainfall: 100 }, { year: 2001, gdp: 1.1, rainfall: 110 }],
items = [...a, ...b],
empty = Object.fromEntries(Object.keys(Object.assign({}, ...items)).map(k => [k, null])),
result = Object
.values(items.reduce((r, o) => {
r[o.year] ??= { ...empty };
Object.assign(r[o.year], o);
return r;
}, {}))
.sort((a, b) => a.year - b.year);
console.log(result);
.as-console-wrapper { max-height: 100% !important; top: 0; }