Sunday, August 25, 2013

#notsobigdata- an epiphany of sorts

I recall writing recently that I was going to lay off of the data-focused posts for a bit and get back to the format of life in and around the Secret Underground Lair for a bit. I was even ready to go with a new topic, a music topic that is near and dear to my heart: why did (fill in the name of a popular and talented musical artist- singer or band) record that atrocity? Furthermore, why did it sell and why did it receive massive airplay? Alas, that will have to wait for another day, because I had one of those moments, and now the Secret Underground Lair's Data Vault will never be the same.

As I had briefly mentioned in an earlier blog, I had updated a table which had the effect of blowing out all of my queries. This is not important right now, as the total amount of data is still under 5Mb(!) and there are still fewer than 900 records in the main data table. For the purpose of updating this table, though, I decided to make a query which would simply list movie titles and formats. So, I sat down with the query wizard and proceeded to make a simple query which would output an alpha list of only movies, as well as their formats.

Simple, right? Muahaha, I guess the air in the Secret Underground Lair was a bit thin when I tried to do this, because I kept on trying the same thing and I kept on getting the same error. Suddenly, as I was staring at the Criteria in Design View, it hit me: it was not working because when I tried to run with the criterion I had selected, the criterion was not recognized. In plain English, I wanted what I thought was data, but Access interpreted it as a displayed name.

Since there are only two pieces of data in the table, I figured out that Access wanted to see the record's autonumber rather than the name displayed in the target table. I switched the so that Criteria==9 (where 9==the record number of the piece of data that I actually wanted to use), and it worked. Which means, if I'm feeling adventurous, I'll save a copy of the database and figure out how the relationships are set up, and then correct so they are pointing to the desired field in the source table. Then, I will eliminate the autonumber field, setting up the only remaining data as the primary key.This will lighten the database a bit, and make it much easier to write queries, as the criteria will all be in plain English. And going forward, all new single field tables will not have an autonumber.

Next time, I'll try to get those songs out, as well as an idea for a relatively ultralight database that I had some ideas about. Until then,...

As always, I am hochspeyer, blogging data analysis and management so you don't have to.

No comments:

Post a Comment