Josep Ferrer Profile picture
Dec 27, 2023 โ€ข 14 tweets โ€ข 5 min read โ€ข Read on X
Ever felt confused by SQL's execution flow? ๐Ÿค”

Then you better stay with me!

Today let's exemplify SQL's execution order with a simple query๐Ÿ‘‡๐Ÿป Image
1๏ธโƒฃ ๐—ฆ๐—ง๐—”๐—ฅ๐—ง๐—œ๐—ก๐—š ๐—™๐—ฅ๐—ข๐—  ๐—ข๐—จ๐—ฅ ๐—ฅ๐—”๐—ช ๐—ง๐—”๐—•๐—Ÿ๐—˜
We use a dummy table with the salary of employees depending on their field and experience,

๐ŸŽฏ Our main goal?
Understand the field that earns the most. Image
2๏ธโƒฃ ๐—ฆ๐—ค๐—Ÿ ๐—ค๐—จ๐—˜๐—ฅ๐—ฌ ๐—ฆ๐—ง๐—ฅ๐—จ๐—–๐—ง๐—จ๐—ฅ๐—˜ (to use)
We define a query to obtain our goal data. Image
3๏ธโƒฃ ๐—ฆ๐—ค๐—Ÿ ๐—”๐—ฆ ๐—” ๐——๐—˜๐—–๐—Ÿ๐—”๐—ฅ๐—”๐—ง๐—œ๐—ฉ๐—˜ ๐—Ÿ๐—”๐—ก๐—š๐—จ๐—”๐—š๐—˜
SQL expects statements to be written in a specific orderโ€Š...

but their evaluation sequence differs. Image
๐—ฆ๐—ง๐—˜๐—ฃ 1 - ๐—™๐—ฅ๐—ข๐— 
First SQL determines the table where to take the data from.

In our case our dummy table salaries with all the info it contains. Image
๐—ฆ๐—ง๐—˜๐—ฃ 2 - ๐—ช๐—›๐—˜๐—ฅ๐—˜
Here data is filtered.

This is where we specify conditions to narrow down our results.

Think of it as our detective's magnifying glass! ๐Ÿ” Image
๐—ฆ๐—ง๐—˜๐—ฃ 3 - ๐—š๐—ฅ๐—ข๐—จ๐—ฃ ๐—•๐—ฌ
Grouping time!

This step clusters similar data together based on a specified column.

Perfect for summarizing data like employee fields! ๐Ÿ“Š Image
๐—ฆ๐—ง๐—˜๐—ฃ 4 - ๐—›๐—”๐—ฉ๐—œ๐—ก๐—š
It's like WHERE's big sibling.

HAVING filters groups, especially useful after GROUP BY.

Only the groups meeting our criteria will pass this gate. Image
๐—ฆ๐—ง๐—˜๐—ฃ 5 - ๐—ฆ๐—˜๐—Ÿ๐—˜๐—–๐—ง
The star of the show!

Here we pick the columns we want to see in our final output.

Like an artist choosing colors for their painting ๐ŸŽจ Image
๐—ฆ๐—ง๐—˜๐—ฃ 6 - ๐—ข๐—ฅ๐——๐—˜๐—ฅ ๐—•๐—ฌ
Now we arrange our results.

This is where we decide the sorting order, like arranging books on a shelf. ๐Ÿ“š Image
๐—ฆ๐—ง๐—˜๐—ฃ 7 - ๐—Ÿ๐—œ๐— ๐—œ๐—ง
The final touch!

Here we can limit the number of rows in our result set.

It's like picking the top 10 hits of the year. ๐ŸŒŸ Image
4๏ธโƒฃ ๐—™๐—œ๐—ก๐—”๐—Ÿ ๐—ฅ๐—˜๐—ฆ๐—จ๐—Ÿ๐—ง
And voilร ....!

we have our neatly arranged and insightful data, ready for action!

Next week we will see how a JOIN command would affect this flow! Image
Still no clear how SQL's execution flow works?

Check my previous thread to further understand it! ๐Ÿ‘‡๐Ÿป
And that's all for now

If you liked this thread, I am sharing Data Science and AI content.
So don't forget to follow me to get more content like this! (@rfeers)

RT the tweet below to help me share the word! :D

โ€ข โ€ข โ€ข

Missing some Tweet in this thread? You can try to force a refresh
ใ€€

Keep Current with Josep Ferrer

Josep Ferrer Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @rfeers

Sep 15, 2025
How to make your LLMs smarter and more efficient explained!๐Ÿ‘‡๐Ÿป

(Don't forget to bookmark for later ๐Ÿ˜‰) Image
Creating an LLM demo is a breeze.
But... refining it for production? That's where the real challenge begins! ๐Ÿ› ๏ธ

Teams often grapple with LLMs lacking deep knowledge or delivering inaccurate outputs.

How do we fix this?
Optimization isn't a one-size-fits-all. Approach it along two axes:

๐Ÿง  ๐—–๐—ผ๐—ป๐˜๐—ฒ๐˜…๐˜ ๐—ข๐—ฝ๐˜๐—ถ๐—บ๐—ถ๐˜‡๐—ฎ๐˜๐—ถ๐—ผ๐—ป: Is the model missing the right info?
โš™๏ธ ๐—Ÿ๐—Ÿ๐—  ๐—ข๐—ฝ๐˜๐—ถ๐—บ๐—ถ๐˜‡๐—ฎ๐˜๐—ถ๐—ผ๐—ป: Is the model's output off-target? ๐ŸŽฏ

Let's break down the three primary tools ๐Ÿ’ฅ
Read 10 tweets
Apr 19, 2025
Multiple-class Logistic Regression clearly explained ๐Ÿ‘‡๐Ÿป

(Don't forget to bookmark for later! ๐Ÿ˜‰) Image
By default, Logistic Regression is like a coin toss - heads or tails, A or B.

But what if you have multiple classes?

That's where we adapt our model for MULTIPLE CHOICES!
There are two main ways:
1๏ธโƒฃ ๐—ข๐—ก๐—˜-๐—ฉ๐—ฆ-๐—ฅ๐—˜๐—ฆ๐—ง (๐—ข๐˜ƒ๐—ฅ):
The Logistic Regression model excels in classifying binary choices.

So... what if we train multiple Logistic Regression classifiers for every class?

๐Ÿ’ก The idea would be to focus on classifying a single class vs the rest. Image
Read 13 tweets
Apr 15, 2025
Simple Linear Regression exemplified for dummies๐Ÿ‘‡๐Ÿป

(Don't forget to bookmark for later! ๐Ÿ˜‰) Image
1๏ธโƒฃ ๐——๐—”๐—ง๐—” ๐—š๐—”๐—ง๐—›๐—˜๐—ฅ๐—œ๐—ก๐—š ๐—ฃ๐—›๐—”๐—ฆ๐—˜
We're using height and weight - a classic duo often assumed to have a linear relationship.

But assumptions in data science? No way! ๐Ÿง

Let's find out:
- Do height and weight really share a linear bond? Image
Do you like this post?

Then join my DataBites newsletter to get all my content right to your mail every Sunday! ๐Ÿงฉ

๐Ÿ‘‰๐Ÿป ๐Ÿค“databites.tech
Read 18 tweets
Apr 14, 2025
Linear Regression clearly explained ๐Ÿ‘‡๐Ÿป Image
Linear regression is the simplest statistical regression method used for predictive analysis.

It can be performed with multiple variables.... but today we'll focus on a single one.

Also known as Simple Linear Regression.
1๏ธโƒฃ ๐—ฆ๐—œ๐— ๐—ฃ๐—Ÿ๐—˜ ๐—Ÿ๐—œ๐—ก๐—˜๐—”๐—ฅ ๐—ฅ๐—˜๐—š๐—ฅ๐—˜๐—ฆ๐—ฆ๐—œ๐—ข๐—ก
In Simple Linear Regression, we use one independent variable to predict a dependent one.

The main goal? ๐ŸŽฏ
Finding a line of best fit.

It's simple yet powerful, revealing hidden trends in data. Image
Read 13 tweets
Mar 15, 2025
Linear Regression clearly explained ๐Ÿ‘‡๐Ÿป Image
Linear regression is the simplest statistical regression method used for predictive analysis.

It can be performed with multiple variables.... but today we'll focus on a single one.

Also known as Simple Linear Regression.
1๏ธโƒฃ ๐—ฆ๐—œ๐— ๐—ฃ๐—Ÿ๐—˜ ๐—Ÿ๐—œ๐—ก๐—˜๐—”๐—ฅ ๐—ฅ๐—˜๐—š๐—ฅ๐—˜๐—ฆ๐—ฆ๐—œ๐—ข๐—ก
In Simple Linear Regression, we use one independent variable to predict a dependent one.

The main goal? ๐ŸŽฏ
Finding a line of best fit.

It's simple yet powerful, revealing hidden trends in data. Image
Read 13 tweets
Mar 13, 2025
The Transformer's encoder clearly explained ๐Ÿ‘‡๐Ÿป Image
1๏ธโƒฃ ๐—ช๐—›๐—”๐—ง'๐—ฆ ๐—ง๐—›๐—˜ ๐—˜๐—ก๐—–๐—ข๐——๐—˜๐—ฅ? ๐Ÿง 

The Encoder is the part responsible for processing input tokens through self-attention and feed-forward layers to generate context-aware representations.

๐Ÿ‘‰ Itโ€™s the powerhouse behind understanding sequences in NLP models. Image
Are you enjoying this post?

Then join my newsletter DataBites to get all my content right to your mail every week! ๐Ÿงฉ

๐Ÿ‘‰๐Ÿป databites.tech
Read 16 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(