Reputation: 12092
I can't seem to word this properly by I need to group by date
and foo
:
// Using date as numbers for clear example
// date key should be grouped into a key
// Under that key (date), I should see another key of "foo"
let list = [
{ date: "1", foo: "me", result: "meFoo"},
{ date: "1", foo: "me2", result: "meYou"},
{ date: "1", foo: "me3", result: "meHe"},
{ date: "2", foo: "me", result: "meHim"},
{ date: "2", foo: "me2", result: "meHim"}
]
let grouped = _.groupBy(list, "date") // keys ie 1, 2
// I need another key under the date key, possible?
let grouped = _.groupBy(list, "date", (val) => _.groupBy(val), "foo") // result not expected
I'm getting a bit complex here so I used underscore to ease the complexity. Is this possible?
I'll then do something like, I could use _.pluck
:
Object.keys(grouped).map( dateKey => Object.keys(grouped[dateKey]).map(fooKey => console.log( grouped[dateKey][fookey][0] )) )
That really looks wrong on the highest level.
Expected results to look something like:
DATE // key
|
|___ FOO // key
|
|__ array foo objects
Upvotes: 0
Views: 1987
Reputation: 4654
list.reduce((acc, item) => {
const key1 = item.date
const group1 = acc[key1] || {}
const key2 = item.foo
const group2 = group1[key2] || []
return {
...acc,
[key1]: {
...group1,
[key2]: group2.concat(item)
}
}
}, {})
let grouped = {}
list.forEach((item) => {
// Create an empty object if this key has not already been set:
grouped[item.date] = grouped[item.date] || {}
// Create an empty list if this key has not already been set:
grouped[item.date][item.foo] = grouped[item.date][item.foo] || []
// Push the item to our group:
grouped[item.date][item.foo].push(item)
})
In python this would be:
from collections import defaultdict
_list = [
{"date":"1","foo":"me","result":"meFoo"},
{"date":"1","foo":"me2","result":"meYou"},
{"date":"1","foo":"me3","result":"meHe"},
{"date":"2","foo":"me","result":"meHim"},
{"date":"2","foo":"me2","result":"meHim"}
]
grouped = defaultdict(lambda: defaultdict(list))
for item in _list:
grouped[item['date']][item['foo']].append(item)
Upvotes: 5
Reputation: 27986
You were nearly there. For the second level of grouping you need to map across the values of the first grouping:
let result = _.chain(list)
.groupBy('date')
.mapObject( grp => _.groupBy(grp, 'foo'))
.value();
Upvotes: 1
Reputation: 5280
This will give you the desired array.
_.map(_.groupBy(list, 'date'), (value, key) => ({[key]: _.groupBy(value, 'foo')}))
I hope you want such result:
var list = [{ date: "1", foo: "me", result: "meFoo" }, { date: "1", foo: "me2", result: "meYou" }, { date: "1", foo: "me3", result: "meHe" }, { date: "2", foo: "me", result: "meHim" }, { date: "2", foo: "me2", result: "meHim" }],
grouped = _.map(_.groupBy(list, 'date'), (value, key) => ({[key]: _.groupBy(value, 'foo')}));
console.log(grouped);
.as-console-wrapper { max-height: 100% !important; top: 0; }
<script src="https://cdnjs.cloudflare.com/ajax/libs/underscore.js/1.8.3/underscore-min.js"></script>
Upvotes: 2
Reputation: 6546
Here is another way to achieve this using function
composition. I have used Ramda library for this. First, group it via date and then group it via foo. Also, please note that I didn't find any curried function for iterating object keys, that is why I added it myself.
let list = [
{ date: "1", foo: "me", result: "meFoo"},
{ date: "1", foo: "me2", result: "meYou"},
{ date: "1", foo: "me3", result: "meHe"},
{ date: "2", foo: "me", result: "meHim"},
{ date: "2", foo: "me2", result: "meHim"},
{ date: "2", foo: "me2", result: "meHimAnother"}
]
function iterKeys(fn) {
return function(obj) {
for (var key in obj) {
obj[key] = fn(obj[key])
}
return obj;
}
}
var group = R.compose(iterKeys(R.groupBy(R.prop('foo'))), R.groupBy(R.prop('date')))
console.log(group(list));
<script src="https://cdnjs.cloudflare.com/ajax/libs/ramda/0.25.0/ramda.js"></script>
Upvotes: 0