Fred Easey
@fred-easey.bsky.social
Lead Analyst with 10+ years experience working in the mobile games industry.
Tidyverse Fan, Norfolk Terrier Wrangler and part-time bass player.
Tidyverse Fan, Norfolk Terrier Wrangler and part-time bass player.
@twitmericks.bsky.social hi Mick. I’m trying to sign up for the Times weekly puzzle newsletter, but I can’t for the life of me figure out how to sign up. Can you point me in the right direction?
September 16, 2025 at 9:53 AM
@twitmericks.bsky.social hi Mick. I’m trying to sign up for the Times weekly puzzle newsletter, but I can’t for the life of me figure out how to sign up. Can you point me in the right direction?
I've been using lets-plot a fair bit recently (migrating from R to Python for analysis work). It does most of thing things I want although the error messages are basically non-existent, so it's hard to debug.
What other packages are folks using? Any love for plotnine or plotly?
What other packages are folks using? Any love for plotnine or plotly?
November 28, 2024 at 4:22 PM
I've been using lets-plot a fair bit recently (migrating from R to Python for analysis work). It does most of thing things I want although the error messages are basically non-existent, so it's hard to debug.
What other packages are folks using? Any love for plotnine or plotly?
What other packages are folks using? Any love for plotnine or plotly?
TIL!
✨Special late night bonus thread✨
The image on the cover of the Joy Division album "Unknown Pleasures" is from a plot that radio astronomer Harold Craft made for his PhD dissertation, using data collected at Arecibo while studying the pulsar discovered by Jocelyn Bell Burnell. (13/12)
The image on the cover of the Joy Division album "Unknown Pleasures" is from a plot that radio astronomer Harold Craft made for his PhD dissertation, using data collected at Arecibo while studying the pulsar discovered by Jocelyn Bell Burnell. (13/12)
November 28, 2024 at 4:19 PM
TIL!
As a long time z-test guy, I’m happy to simplify things by just using t-tests going forwards
www.linkedin.com/feed/update/...
www.linkedin.com/feed/update/...
Ron Kohavi on LinkedIn: #abtesting #abtests #pvalue #t #z | 23 comments
Question: When should we use a Z-test vs. a t-test in #abtesting?
Me: It does not matter for proper A/B tests
Here is a spreadsheet with a few realistic… | 23 comments on LinkedIn
www.linkedin.com
November 26, 2024 at 2:58 AM
As a long time z-test guy, I’m happy to simplify things by just using t-tests going forwards
www.linkedin.com/feed/update/...
www.linkedin.com/feed/update/...
I've found the folks in the Polars Discord to be very helpful with tech questions (I find chatGPT struggles a bit with Polars, perhaps it's too new for the training corpus to cover well)
discord.com/invite/4UfP5...
discord.com/invite/4UfP5...
Join the polars Discord Server!
Check out the polars community on Discord - hang out with 4607 other members and enjoy free voice and text chat.
discord.com
November 12, 2024 at 3:00 PM
I've found the folks in the Polars Discord to be very helpful with tech questions (I find chatGPT struggles a bit with Polars, perhaps it's too new for the training corpus to cover well)
discord.com/invite/4UfP5...
discord.com/invite/4UfP5...
slowly figuring out how to do dplyr type operations in polars
Today: how to get revenue-per-user from a df with a bunch of different ..._revenue cols and a users col:
ts_data_pdf.with_columns(
(cs.ends_with("_revenue") / pl.col("users") * 100).name.suffix("_per_user")
)
Today: how to get revenue-per-user from a df with a bunch of different ..._revenue cols and a users col:
ts_data_pdf.with_columns(
(cs.ends_with("_revenue") / pl.col("users") * 100).name.suffix("_per_user")
)
November 12, 2024 at 2:58 PM
slowly figuring out how to do dplyr type operations in polars
Today: how to get revenue-per-user from a df with a bunch of different ..._revenue cols and a users col:
ts_data_pdf.with_columns(
(cs.ends_with("_revenue") / pl.col("users") * 100).name.suffix("_per_user")
)
Today: how to get revenue-per-user from a df with a bunch of different ..._revenue cols and a users col:
ts_data_pdf.with_columns(
(cs.ends_with("_revenue") / pl.col("users") * 100).name.suffix("_per_user")
)
I'm a Tidyverse fan, but a new role means I'm on Python now.
I've been learning Polars, lets-plot and pingouin (inspired by @emilyriederer.bsky.social python rgonomics series, feels v. close to the tidyverse setup), but should I be embracing pandas/matplotlib/scikit/statsmodels "standard" instead?
I've been learning Polars, lets-plot and pingouin (inspired by @emilyriederer.bsky.social python rgonomics series, feels v. close to the tidyverse setup), but should I be embracing pandas/matplotlib/scikit/statsmodels "standard" instead?
October 29, 2024 at 9:54 AM
I'm a Tidyverse fan, but a new role means I'm on Python now.
I've been learning Polars, lets-plot and pingouin (inspired by @emilyriederer.bsky.social python rgonomics series, feels v. close to the tidyverse setup), but should I be embracing pandas/matplotlib/scikit/statsmodels "standard" instead?
I've been learning Polars, lets-plot and pingouin (inspired by @emilyriederer.bsky.social python rgonomics series, feels v. close to the tidyverse setup), but should I be embracing pandas/matplotlib/scikit/statsmodels "standard" instead?
tempted to try and make a starter pack for data folks working in the games industry 🤔
October 28, 2024 at 5:26 PM
tempted to try and make a starter pack for data folks working in the games industry 🤔
Reposted by Fred Easey
Check out the new `pastum` extension for Positron (and VS Code I guess)—it's {datapasta} for Positron! It works with both R (base R, tibble, and data.frame) and Python (pandas and polars) github.com/atsyplenkov/... #rstats #python
October 28, 2024 at 3:18 PM
Check out the new `pastum` extension for Positron (and VS Code I guess)—it's {datapasta} for Positron! It works with both R (base R, tibble, and data.frame) and Python (pandas and polars) github.com/atsyplenkov/... #rstats #python
I've been using Positron a bit, I like what I see so far, interested to see where it goes 🐍
We've got a brand new, baby website for Positron! Take a look if you are interested in getting started, and please let us know how it goes:
positron.posit.co
positron.posit.co
Positron
A next-generation data science IDE
positron.posit.co
October 28, 2024 at 5:23 PM
I've been using Positron a bit, I like what I see so far, interested to see where it goes 🐍
I'm liking the starter packs feature here, great to see so many familiar names from the Twitter datasci community here already 💪
October 28, 2024 at 2:40 PM
I'm liking the starter packs feature here, great to see so many familiar names from the Twitter datasci community here already 💪