On a related note, Mode's SQL tutorial is the best comprehensive introduction to the language that I've come across. If you go through all of it you can get from zero knowledge to 95% of the SQL skills you'll need as a data analyst.
I am trying to learn intermediate to advanced SQL as my employer wants me to but apart from the the most basic keywords I don't see a lot of difference between Basic to Intermediate SQL. Or maybe by intermediate it means designing a good data model which shouldn't be concerned with SQL. My question is, if someone asks me to learn the advanced SQL constructs, where do I go? What do I look for?
I found that the book SQL Antipatterns helped me figure out some advanced SQL constructs. It presents a problem, shows the naive solution, takes apart that solution, and shows a better way.
Impressive tutorial! their advanced level tutorial is pretty awesome and helps someone who wants to brush up SQL skills before having analyst interviews. They even have Python tutorial too which I want to try as well.
This can't be overstated enough. Finding people who understand SQL is difficult enough, but finding people who intuitively "get" data is nearly impossible. And lately anyone who completed a few tensorflow tutorials thinks they're a Data Scientist and every single data question can be answered accordingly. Reminds me of the hadoop craze a decade ago.
If you're looking for something free, open source, selfhosted, or doesn't require programming knowledge (but allows SQL if needed), there's Metabase: https://metabase.com/
Thank you for your work! Metabase has been a tremendous help in getting my coworkers interested in data and analytics, and is probably the sole reason why our call center is finally registering data in a structured and organized way.
Metabase is great but try as I might, I cannot get it to connect to a Postgres database over SSL. So... I'm using Redash instead, where it was easier to make that happen.
This blog post asserts that Mode Studio supports R Notebooks. It's worth noting that these don't appear to have anything to do with the R Notebooks that you get in RStudio[0].
They appear to be identical to its Python notebooks except for the language, like running a Jupyter with an R kernel.
Somewhat related, but I absolutely love Mode. In a previous role my team used it extensively - by far the easiest tool of its kind I've used. Great support too.
Can't recommend their tutorials enough, to echo another here they really cover everything you need to get started
"The public function is available free of charge for public use and allows you only to view information other Mode users disclose using the public function. When you use the public function, you may disclose or upload Your Content using only the public function. "
Hey! I'm Mode's CEO. We host a database that contains all the data for tutorials. You can also upload CSVs to this database and do analysis to share with others, but the data and analysis are publicly visible. This is a feature we have had for awhile.
The paragraph below the one you cited says "The private function is available in two versions: (i) free of charge (with limited functionality) and (ii) as a premium paid Service (“Premium Service”)." This is the thing that's new today -- you can now work with private data for free.
Regardless of whether you are on a free or paid account, you have two options:
-Public: Upload data or use data that is already in a database that Mode hosts. All data and analysis done here is visible to anyone. This is most commonly used with the SQL tutorials that others have posted about here.
-Private: Connect your database to Mode and do analysis on your own data. This is only visible to people you add to your Mode Organization. In fact, making your private data accessible to people outside your Organization is a paid feature and comes with more granular permission controls.
We have tried to make this as clear as possible in the product and ToS, but looks like we missed on the latter. We will revise to make it clearer.
None of these platforms "get" your data other than running the queries in your databases and displaying the results (and sometimes caching it for future lookups).
It runs in your browser which is key to allowing people across the organization easy access to data. Its great! I was one of their early customers and I still use it most days.
I adore Mode - i find it the best SQL editor and dashboard by a huge margin. I mostly just use it as a SQL editor and then there’s an amazing ability to turn that SQL into exceptionally useful dashboards.
I can see the use case of this being similar to having Excel connected to an SQL server and creating some pivot charts off of the connection. Then rather than emailing the Excel file (which might be quite large depending on how much data is pulled into Excel), emailing a URL which is much cleaner.
Am I correct in thinking that Mode is not really a reporting platform (e.g., QlikView, Power BI) and more of an investigative, analytics-sharing tool?
Yes, it's more exploratory and lower-level. It can do the usual "dashboard" style reports but its closer to Juypter notebooks and allows you to run python + sql (+ R now) and even edit the HTML to make your own visualizations.
This look amazing, and the WYSIWYG editor mode is what R Markdown notebooks (in RStudio) are missing right now.
However, there seems to be no way of uploading a private .csv? I need to make it public OR allow access to my company database (, which is simply not going to happen unless you have an on-premises solution)?
I have to sign up to see this in action? No video demo or anything? I don't even know if this is a software where you download or if it's a web service cloud thing.
I think this is such a bad way to announce a product. Or lazy way.
Mode is a great tool for running a query and executing some code afterwards and even distributing that report to others. However, its important to keep in mind that it doesn't have any real organizational tools which limit its use case. There are no tools for versioning and sharing notebooks (there is actually a shared query feature that is available at extra cost) and code (and you can only run the versions of code they have installed), or modeling your data and report parameterization is very difficult. Also, (perhaps due to being solely browser based) there can be issues working with larger data sets.
https://community.modeanalytics.com/sql/tutorial/introductio...