Cumulative distinct count of multiple fields in the same table

Question

I have a property database and I am currently designing a query for additional characteristics of the properties. In this case the two tables in question are those:

This is the first Characteristics table which is basically where I store the meta for each characteristic.

This is the CharacteristicsPerProperty database where I store the relation and final value of the specific characteristic.

So basically what I want to do is count all the unique id_propiedad ( property_id ) so that I can show how many I have available with that specific characteristic. The query is the following

SELECT 
    COUNT(DISTINCT(CaracteristicasPorPropiedad.id_propiedad)) AS disponibles
FROM
    CaracteristicasPorPropiedad
WHERE
    CaracteristicasPorPropiedad.id_caracteristica = 1
    AND
    CaracteristicasPorPropiedad.valor = 3

So basically the output is something like:

| disponibles |
|=============|
| 264         |

Everything works as expected up to here, but I can also include another parameter, let's say in the previous example I included Recámaras (Bedrooms), but I also want to include Baños (Bathrooms). I tried

SELECT 
    COUNT(DISTINCT(CaracteristicasPorPropiedad.id_propiedad))
FROM
    CaracteristicasPorPropiedad
WHERE
    (CaracteristicasPorPropiedad.id_caracteristica = 1 AND CaracteristicasPorPropiedad.valor = 3)
    AND
    (CaracteristicasPorPropiedad.id_caracteristica = 2 AND CaracteristicasPorPropiedad.valor = 2.5)

The thing is that this query returns 0.

If I switch the AND to an OR it returns 282 results. I basically need the combination of both. I am a little stuck with this. What could be a query to mix multiple levels of this? It can be used with all 10 available dimensions for this characteristic.

Answer 1

You can get the id_propiedad using aggregation:

SELECT cpd.id_propiedad
FROM CaracteristicasPorPropiedad cpd
WHERE (cpd.id_caracteristica = 1 AND cpd.valor = 3) OR
      (cpd.id_caracteristica = 2 AND cpd.valor = 2.5)
GROUP BY cpd.id_propiedad
HAVING COUNT(DISTINCT cpd.id_caracteristica) = 2;

If you want the count, use a subquery:

SELECT COUNT(*)
FROM (SELECT cpd.id_propiedad
      FROM CaracteristicasPorPropiedad cpd
      WHERE (cpd.id_caracteristica = 1 AND cpd.valor = 3) OR
            (cpd.id_caracteristica = 2 AND cpd.valor = 2.5)
      GROUP BY cpd.id_propiedad
      HAVING COUNT(DISTINCT cpd.id_caracteristica) = 2
     ) p;