Databases
StackOverflow2013
Postsdbo
POCouchDB Modelling - Time filtered and group data
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POCouchDB Modelling - Time filtered and group data
    primarykey
    data
    text
    <p>I'm trying to build up my understanding of the CouchDB and how to model data for some real world scenarios. I've done as many 'get me blog posts by date' as I can for now ;)</p> <p>Given documents like so:</p> <pre><code>{ "_id": "couch1", "_rev": "2-338d0a592ad1e5570000002b00000000", "eventType": "event1", "date": 1328805860000 } { "_id": "couch2", "_rev": "1-1e0315c2e1ca7f5f0000002b00000000", "eventType": "event1", "date": 1328133600000 } { "_id": "couch3", "_rev": "1-154cd416b78cb2ef0000002b00000000", "eventType": "event2", "date": 1325434920000 } </code></pre> <p><em>Where the date is an epoch would it be possible to ask Couch to make a view where you asked for all "events" that happened betweem two timestamps and then group that data by the "eventType"?</em></p> <p>So using the above and assuming the timestamps passed in encompass those documents - we'd want to see output:</p> <pre><code>"event1": 2 "event2": 1 </code></pre> <p><em>Further info I have obtained</em> </p> <p>I'm aware that Couch will sort by key so if I wanted a 'top 10' then that would be a second phase but I can handle that.</p> <p>So the core problem here is that you are filtering by one column but then grouping by another?</p> <p>If we use the following map function:</p> <pre><code>function (doc) { emit([doc.date, doc.eventType], doc.eventType); } </code></pre> <p>with a <code>count</code> reduce function we see that because the timestamps are essentially unique Couch cannot group and key has the value 1.</p> <p>So you can change the map function to the following:</p> <pre><code>function (doc) { emit([doc.eventType, doc.date], doc.eventType); } </code></pre> <p>And then change the <code>group level</code> to 1 which will group correctly by event but your data cannot then be sliced by time because your primary ordering is by the event name, meaning that time ordering is now broken?</p> <p>Do people have any war stories on this? Does this need to be done with re-reduce?</p> <p>Many thanks in advance to anyone taking time to read this</p> <p>Eggsy</p>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. This table or related slice is empty.
    1. USeggsy84
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    3. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POCouchDB Modelling - Time filtered and group data
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POCouchDB Modelling - Time filtered and group data
      1. This table or related slice is empty.
      1. VTUpMod
    1. COThe answers are great; however notice that sanning/selecting by key **A** and then grouping by key **B** is fundamentally a two-dimensional query; and CouchDB views are *always* one-dimension.
      singulars
      1. POCouchDB Modelling - Time filtered and group data
      1. USJasonSmith
    2. COSo we're saying essentially it can't be done in the most efficient manner and drive it on to the DB. We can always do the grouping on the client but we'd rather not. It seems that this kind of problem would be a typical form of analytics problem and I wonder if any other NoSQL solutions that are built for large scale data could perform such a task
      singulars
      1. POCouchDB Modelling - Time filtered and group data
      1. USeggsy84
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload