How would this query in mongodb - mongodb

Question

example document
{
"_id" : ObjectId("5338796453370917f05bb064"),
"Sigla" : "CE",
"Regiao" : "Nordeste",
"Codigo" : 2306009,
"Municipio" : "Iracema",
"1991" : 52.40499877929688,
"2000" : 108.7089996337891,
"IDHEducacao" : {
"1991" : 0.516,
"2000" : 0.735
}
}
{
"_id" : ObjectId("5338796453370917f05bb065"),
"Sigla" : "CE",
"Regiao" : "Nordeste",
"Codigo" : 2306108,
"Municipio" : "Irauçuba",
"1991" : 47.72299957275391,
"2000" : 62.65800094604492,
"IDHEducacao" : {
"1991" : 0.491,
"2000" : 0.692
}
}
---> Mongodb
I made the following query
{"$group":
{
"_id":{"Regiao":"$Regiao"},
"IDHEducao_max_2000" : {"$max" : "$2000"},
}
}
I want to show the region, the largest index of the field in 2000, and what is the municipality that owns this index. But I'm not getting

Answer 1

Looks like 2000 is the name of one of the fields in your document, which I find strange.
The SQL below:
SELECT Regiao, MAX( 2000 ) AS Indice FROM table1 GROUP BY Regiao
can be written in MongoDB as
db.table1.aggregate([
{"$group": {
"_id":{"Regiao":"$Regiao"},
"IDHEducao_max_2000" : {"$max" : "$2000"}}
},
{"$project": {"_id":0, "Regiao":"$_id", "Indice":"$IDHEducao_max_2000"}}
]}
But this SQL:
SELECT Sigla, Regiao, Municipio, MAX( 2000 ) AS Indice FROM table1 GROUP BY Regiao
is NOT valid. When you use GROUP BY, you can only select fields used in the GROUP BY or aggregated values of other fields (i.e., SUM(), COUNT(), etc..). However, if all you need is some value for the other fields, you could use the $first or $last operators. Note that these operators are usually used only after a sort phase to get min/max:
db.table1.aggregate([
{"$group": {
"_id":{"Regiao":"$Regiao"},
"Sigla" : {"$first" : "$Sigla"}}
"Municipio" : {"$first" : "$Municipio"}}
"IDHEducao_max_2000" : {"$max" : "$2000"}}
},
{"$project": {"_id":0, "Sigla":1, "Regiao":"$_id", "Municipio":1,
"Indice":"$IDHEducao_max_2000"}}
]}
EDIT: OP was updated with the question below:
I want to show the region, the largest index of the field in 2000, and what is the municipality that owns this index.
If you use the $sort phase of the aggregation pipeline followed by $group phase and make use of the $first operator, you can get the results you want:
db.table1.aggregate([
// Sort by (City ASC, Index DESC)
{$sort:{"Regiao":1, "2000":-1}},
// Group by City and get the max Index and corresponding Municipality
{$group:{
_id:"$Regiao",
Index:{$first:"$2000"},
Municipio:{$first:"$Municipio"}}
}
])