Adrian Perea

477 posts

Adrian Perea banner
Adrian Perea

Adrian Perea

@iamperea

Data engineer @thinkdatasci. I geek out about tech sometimes

Makati, Philippines Katılım Ekim 2019
150 Takip Edilen66 Takipçiler
Sabitlenmiş Tweet
Adrian Perea
Adrian Perea@iamperea·
There are many different ways to increase your productivity as a dev. Here are 7 simple ways you can start doing today! #100DaysOfCode #productivity
English
1
3
7
0
ᴄʜʀɪꜱʟʏɴ ABERIN
ᴄʜʀɪꜱʟʏɴ ABERIN@hi_itsme___·
📢 CHRISLYN ABERIN exposing thread? It has been <almost> a year since she borrowed money from me. In case you forgot, @chrslynnnnn_
English
3
1
1
2.2K
Adrian Perea retweetledi
Peer Richelsen
Peer Richelsen@peer_rich·
Remote work was a mistake. After two years of being fully remote, we're now going back to in-office. This is a hard task because we need to relocate every team member. Check out our new headquarter and why we are getting back in person. A thread 🧵
English
213
487
3.9K
2.4M
Adrian Perea retweetledi
Alex Monahan
Alex Monahan@__AlexMonahan__·
@DuckDB now supports upserts through the "on conflict" clause! For now, you just need to install from the latest master branch, but this will be in the next release! #on-conflict-clause" target="_blank" rel="nofollow noopener">duckdb.org/docs/sql/state…
English
2
3
24
1.1K
Adrian Perea retweetledi
Jay Henderson
Jay Henderson@jay_henderson·
If you're a data nerd, you won't want to miss this article on the shift from the modern data stack to the modern data graph. By @stkbailey. bit.ly/3hLEMrD
English
0
2
6
379
Adrian Perea retweetledi
Andrew Gazdecki
Andrew Gazdecki@agazdecki·
@elonmusk @paulg Can you free @acquire for me? 🙏 It would truly make my year. It’s been unused and suspended. No worries if not just shooting my shot.
Andrew Gazdecki@agazdecki

Hey @elonmusk got Russ Hanneman to weigh in on the suspended @acquire handle for some laughs. For the love of startups please let us use this Twitter handle instead of @acquiredotcom for our rebrand. Just shooting my shot!

San Mateo, CA 🇺🇸 English
6
6
99
0
Adrian Perea retweetledi
PythonPH
PythonPH@pythonph·
PyCon PH is back! 🎉 It's been almost 4 years since we hosted PyCon APAC 2019 and went on a strategic pause. Finally, PyCon Philippines is back with the theme "ugnayan", celebrating the connections we formed and continue to build as a community. Stay tuned and see you next year!
PythonPH tweet media
English
2
26
64
0
Adrian Perea retweetledi
Thinking Machines
Thinking Machines@thinkdatasci·
Friend or foe? Let’s break down the ways that AI can ⚡supercharge⚡ our decisions to improve our daily lives and how we do business. Follow our blog post series, We’re Better with AI, for more content on becoming an AI-driven organization. Full post at thinkingmachin.es/stories/why-an….
English
0
5
4
0
Adrian Perea retweetledi
Alessandro Martinello
Alessandro Martinello@ale_martinello·
Without DuckDB our current lead project would not exist. In a closed environment w/o access to internet (cloud) and only python available we need to analyze ~1B transactions + 128M rows dataset (+ several ~5M rows ones). Once the db is set up, even complex queries take ~15 secs
English
0
5
36
0
Alex Monahan
Alex Monahan@__AlexMonahan__·
@iamperea @duckdb @duckdblabs @motherduckdb DuckDB is MIT licensed open source software owned by a foundation. DuckDBLabs is a consulting and custom development company formed by the creators and many of the maintainers of DuckDB. Motherduck is a venture backed startup using DuckDB tech in the cloud! That help as a start?
English
1
0
5
0
Adrian Perea
Adrian Perea@iamperea·
@hfmuehleisen I watched your CWI talk (youtube.com/watch?v=obgY1D…). For the Spark vs Duck DB tests that you did, how did you get data into the duckdb node? I know Spark can do it via its multiple data connectors. Would like to know what a good practice for duckdb would be
YouTube video
YouTube
English
0
0
0
0
Adrian Perea retweetledi
Paul Graham
Paul Graham@paulg·
Today I talked to a startup doing so well that they had no current problems that needed solving. Profitable, growing ~20x a year (not a typo), only 9 employees. This is so rare that I didn't know what to do. We ended up talking about problems they might have in the future.
English
183
172
5.3K
0
Alex Monahan
Alex Monahan@__AlexMonahan__·
@iamperea @duckdb Howdy! When you export, we save the data off as a parquet file per table, and sql scripts to rebuild your schemas. So it is no longer a single file format! It's slower to import than to use an existing DB, but we use exports if the DuckDB storage format changes between versions.
English
1
0
6
0
Adrian Perea
Adrian Perea@iamperea·
Hi @duckdb! This is driving me nuts: what is the difference between using a persistent database (e.g. in python: duckdb.connect(database="database.duckdb")) vs saving the contents of a DB using EXPORT and then using IMPORT in subsequent sessions?
English
1
1
1
0
Adrian Perea retweetledi
Jacob Matson
Jacob Matson@matsonj·
based on my tweets yesterday, just wrote up how I increased query performance 80x by using DuckDB instead of postgresql. dataduel.co/revisiting-dat…
English
3
16
117
0
Adrian Perea
Adrian Perea@iamperea·
💎 ✋️ 💎 ✋️ 💎 ✋️ This is the closest thing the crypto community has to a prayer
English
0
0
1
0
Adrian Perea retweetledi
Alex Monahan
Alex Monahan@__AlexMonahan__·
The power of @DuckDB and @ApacheArrow: "We can select 304,851 interesting rows from all 1,547,741,381 in the 10 year dataset in < 3 seconds on a laptop!" Great demo!
@[email protected]@GarsBar35Plus

Followed up with the 10 year, 1.5 billion row dataset. Duckdb + Arrow filtered through it with ease on a laptop @streamlit code: github.com/gerardrbentley… (No demo because the data is 40 GB 💾 download instructions via s3 in repo)

English
0
5
47
0