# OECD:MEI JSON Data File with ONLY Quarterly Data Available?

**URL:** <https://dbnomics.discourse.group/t/oecd-mei-json-data-file-with-only-quarterly-data-available/384>\
**Category:** Uncategorized\
**Created:** [November 27, 2020, 6:52pm UTC](https://dbnomics.discourse.group/t/oecd-mei-json-data-file-with-only-quarterly-data-available/384 "2020-11-27T18:52:57Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![MT\_MANC](https://avatars.discourse-cdn.com/v4/letter/m/57b2e6/32.png) [@MT\_MANC](https://dbnomics.discourse.group/u/MT_MANC)\
**Post date:** [November 27, 2020, 6:52pm UTC](https://dbnomics.discourse.group/t/oecd-mei-json-data-file-with-only-quarterly-data-available/384/1 "2020-11-27T18:52:58Z")

</div>

We are accessing OECD: MEI data via Python.:  
[https://git.nomics.world/dbnomics-json-data/oecd-json-data/tree/master/MEI/series.jsonl](https://git.nomics.world/dbnomics-json-data/oecd-json-data/tree/master/MEI/series.jsonl)

We only need to access QUARTERLY data for all MEI series which we are then dumping as a .CSV in Rows x Columns format. The **series.jsonl** file we are accessing is massive as it contains Monthly (and Annual) data too; the Python script has to check every “row” in order to select only the Quarterly periodicity data and it’s taking a very long time (its regularly updated too)  
Question: - do you ever store eg MEI JSON files that contain:  
just QUARTERLY data eg “seriesQ.jsonl” or  
just ANNUAL data eg “seriesA.jsonl” or  
just MONTHLY data eg “seriesM.jsonl”

Accessing seriesQ.jsonl would then be MUCH quicker than the current “grand” series.jsonl  
Is this file seperation feasible for MEI (& other datasets containing Monthly data as well as Quarterly & Annual data?)

Most kind

---

<div class="post-metadata">

**Author:** ![thomasbrand](https://avatars.discourse-cdn.com/v4/letter/t/9e8a1a/32.png) [@thomasbrand](https://dbnomics.discourse.group/u/thomasbrand)\
**Post date:** [December 3, 2020, 1:33pm UTC](https://dbnomics.discourse.group/t/oecd-mei-json-data-file-with-only-quarterly-data-available/384/2 "2020-12-03T13:33:58Z")

</div>

We do not store dataset along specific dimensions for now. And I doubt that we will do so in the near future. Indeed, we are trying to be systematic and I don’t see why we would do it for the `frequency` dimension, and not for the `country` dimension for example.

However, it is possible to make a query only on the quarterly series under python or R, like this :  
`library(rdbnomics)`  
`df <- rdb("OECD", "MEI","...Q")`  
It is quite fast (less than 2 minutes with my computer).

---

<div class="post-metadata">

**Author:** ![cbenz](https://yyz2.discourse-cdn.com/free1/user_avatar/dbnomics.discourse.group/cbenz/32/33_2.png) [@cbenz](https://dbnomics.discourse.group/u/cbenz)\
**Post date:** [December 8, 2020, 1:48pm UTC](https://dbnomics.discourse.group/t/oecd-mei-json-data-file-with-only-quarterly-data-available/384/3 "2020-12-08T13:48:48Z")

</div>

Hi @MT_MANC

I’d like to emphasis the fact that DBnomics is primarily designed to be queried via its HTTP API.

Although it is possible to clone `*-json-data` Git repositories and read data from them, they are more like an internal storage shared between the different components of DBnomics. Furthermore, the storage format is subject to change.