Hello and welcome to the Extracts post!
Explaining extracts is very simple indeed, however there are some implications of using them that we must consider and will go through them here.
So, on the data source page there are a couple of buttons on the top right of the screen under the title “Connection”. The first, and (as long as you are online) default choice is “Live”. This simply means that Tableau has a live connection to your data source and will check back with it whenever you ask it to make a computation. If the data source is on your local hard drive, then the only real benefit to it being a live connection is if you decide to go back to your (let’s say excel) dataset and make a change, Tableau will notice that change and apply it accordingly when you return to your Tableau viz (this may require you to press the refresh button on the top left if it hasn’t checked back in a while). Normally I would leave it on live unless I am using a HUGE dataset, as in this case it would be a drain on Tableau’s computing resources each time it refers back to the source.
An Extract by comparison is a “snapshot” of the data source that is (normally) saved in your tableau repository (local hard drive). This happens automatically if you lose your internet connection so that your work doesn’t just disappear in a puff of smoke.
If, for example, you are working for an international institution, whose live connection to the data is bouncing through multiple countries and servers, you may notice some lag each time you try to run a computation (such as making a graph). In this case it may be worthwhile changing your connection to an Extract. HOWEVER, there are some implications to this, if the data is highly sensitive, such as your client lists or sales data etc, a local copy will be made in your computer (which usually has far less security that a well maintained server). When you finish with the data you can go and delete the file locally (it will ask you where to save it so you should be able to re-find it) and your viz will remain intact, but if you want to make any further changes to the viz after this, you will need to reconnect to the original file location.
Another consideration is whether your dataset is static or subject to live updates. Using the sales or customer data example, you company probably updates that file as more data becomes available, so if you are creating a viz for a presentation next week, the data may no longer be applicable or will become outdated if you used an extract at the time of creation.
If you are searching through your local files, and easy way to see if a dataset is live or an extract is to look at the type of file it has been saved as in Tableau. If you are using a “.hyper” or a “.tde” file, this is an Extract, as is anything that ends in “x” (for instance “.twbx”). By contrast, both “.tds” and “.twb” are live connections that will update when you open them and continue to run connections when being used.
Ok that’s all folks, onto the next one!