baggiponte

9 ms·

61.
by baggiponte 4y ago
I recommend PDM over poetry!
62.
by baggiponte 4y ago
Source?
63.
by baggiponte 4y ago
Incidentally, is Tailscale reliable? I recall VPN services are a bit iffy because they need to store user data so it does not really matter the service they offer - if what you are after is privacy, rather than a way to bypass country-based
64.
New ggplot/Altair like interface for Seaborn now in beta(github.com)
2 points by baggiponte 4y ago|1 comments
65.
by baggiponte 4y ago
This brings together the best features of R's ggplot and Altair's features for Seaborn
66.
How to collapse values of a factor in pandas(cmdlinetips.com)
1 points by baggiponte 4y ago|1 comments
67.
by baggiponte 4y ago
This is an area where pandas is bery far behind the R frameworks...
68.
by baggiponte 4y ago
yeah that's a neat trick. of course you can do that with cmd-shift-5 too
69.
by baggiponte 4y ago
I don’t really get how the notification to 400k people was sent :/
70.
by baggiponte 4y ago
oh hi, sorry for the late reply. indeed you are right; i was typing from my phone and was a bit too hasty. i am not a fan of poetry after reading these; TLDR poetry does not support PEP621 about pyproject.toml either, and i feel this is the
71.
by baggiponte 4y ago
Wow this is 1000x than what’s in the article - thanks.
72.
by baggiponte 4y ago
A bit unrelated, but I have been using pdm ( https://pdm.fming.dev/ ) which is something like Poetry but faster and PEP compliant (!)
73.
by baggiponte 4y ago
I think that this means you don’t have to download a copy of the file locally?
74.
by baggiponte 4y ago
I don’t understand what this implies/enables O.o could somebody be kind enough to explain? Thanks!
75.
by baggiponte 4y ago
That’s fascinating! Totally orthogonal: how does PG store data? Do they use columnar specifications such as Arrow & Parquet?
76.
by baggiponte 4y ago
Perhaps we “just” need Excel and Numbers to support reading .parquet files, and come up with a CLI tool to view them like head/tail/awk? Yes, they still cannot be read as plain text so functionality is quite restricted with respec
77.
by baggiponte 4y ago
I am a total newbie here - you speak of DBs because of the backend that needs to be set up to host the users' app?
78.
by baggiponte 4y ago
Alternatives to Heroku besides Porter?
79.
by baggiponte 5y ago
Definitely would achieve more if he open sourced mathematica
80.
by baggiponte 5y ago
That’s concerning - I am also wondering if iTerm too sends this data…
81.
by baggiponte 5y ago
The paper claims TF and PyT fail to exploit linear algebra optimisation tricks: " for instance, reducing the number of scalar operations by applying the distributive law, and automatically identifying the optimal parenthesization of a
82.
TensorFlow and PyTorch under use linear algebra optimisation(arxiv.org)
1 points by baggiponte 5y ago|1 comments
83.
by baggiponte 5y ago
What is Q?
84.
by baggiponte 5y ago
I’m afraid I’m lost - what’s the relationship between HSLS, CUDA and OpenCL?
85.
by baggiponte 5y ago
I struggle a bit to understand what Substrait wants to accomplish. They want to create a representation of data transformations (eg group by…) and relational algebra at what level? So it’s like a BLAS for relational algebra?
86.
by baggiponte 5y ago
Yeah, that I knew: why Click over Typer? Greater control in creating the CLI?
87.
by baggiponte 5y ago
How do you add a new line? I tried $ rich --print "\n hello world" but it just returned \n hello world Also, where can I find the list of emoji and the correspondent ":description:"?
88.
by baggiponte 5y ago
So you are using Click and not Typer?
89.
by baggiponte 5y ago
That's beautiful! Maybe adding tab completions would be useful? For example: $ rich "hello world" -p -a <TAB> ... Where "..." is all the possible values that can be taken in.
90.
by baggiponte 5y ago
I have been looking for something like this for a long time T.T

More ›