Sharon Machlis

@smachlis.bsky.social

2842 Followers 345 Following

Author of Practical R for Mass Communication & Journalism ❤️ #RStats, writing & walkable/bikeable cities Having fun with GenAI Retired tech journalist Also #SmartHome #HamRadio #VibrantAging #Running #Photography #Crochet #ASL https://machlis.com

Bar chart of weekly #RStats posts (Sat–Fri). The week of Jul 25–Jul 31, 2026 had 310 posts, and the week before had 311.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jul 18–Jul 24, 2026 had 311 posts, and the week before had 154.
@claudeai post on Twitter/X: Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits. Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit. Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity. 10:14 PM · Jul 17, 2026 11.4M Views, 3.7K comments, 10K reposts, 41K likes Additional post: We know this has been frustrating, and we want to give you more certainty about what your plan includes. We are making access standard at 50% usage for the plans that use Fable most intensively. Thank you for your patience over the past several weeks. We're continuing to invest in new capacity and will keep everyone updated as we do.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jul 11–Jul 17, 2026 had 154 posts, and the week before had 186.
Hummingbird in flight near flowers
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jul 4–Jul 10, 2026 had 186 posts, and the week before had 205.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jun 27–Jul 3, 2026 had 205 posts, and the week before had 218.
404 Error Like guarantees of future returns, this page doesn’t exist. Try searching, or go to our homepage.
Screenshot of tournament knockout-stage schedule set to US Eastern time. Tabs read "By team," "By date," "Knockout stage," and "Standings." A button says "Update scores." Round of 32: Sun, Jun 28 · 3:00 PM EDT, South Africa vs Canada; Mon, Jun 29 · 1:00 PM EDT, Brazil vs Japan; Mon, Jun 29 · 4:30 PM EDT, Germany vs Paraguay; Mon, Jun 29 · 9:00 PM EDT, Netherlands vs Morocco; Tue, Jun 30 · 1:00 PM EDT, Ivory Coast vs Norway; Tue, Jun 30 · 5:00 PM EDT, France vs Sweden; Tue, Jun 30 · 9:00 PM EDT, Mexico vs Ecuador; Wed, Jul 1 · 12:00 PM EDT, England vs Congo DR; Wed, Jul 1 · 4:00 PM EDT, Belgium vs Senegal. Round of 16 through final show scheduled dates and times, with all teams listed as "TBD." Alt text by ChatGPT
29 cent US commemorative postage stamps showing a dressing of a player and a soccer ball)
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jun 20–Jun 26, 2026 had 218 posts, and the week before had 211.
"My value as a consultant is not defined by the R code I write. It is defined by the strategic clarity, values, and outcomes I help my clients achieve through the tools & systems I create!"
Big soccer ball sculpture in Boston's Copley Square
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jun 13–Jun 19, 2026 had 211 posts, and the week before had 255.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Jun 6–Jun 12, 2026 had 255 posts, and the week before had 220.
Deer in the woods
Bar chart of weekly #RStats posts (Sat–Fri). The week of May 30–Jun 5, 2026 had 220 posts, and the week before had 234.
Bar chart of weekly #RStats posts (Sat–Fri). The week of May 23–May 29, 2026 had 234 posts, and the week before had 256.
Screenshot of a weather station reading showing 53.6 degrees.
Bar chart of weekly #RStats posts (Sat–Fri). The week of May 16–May 22, 2026 had 256 posts, and the week before had 227.
Bar chart of weekly #RStats posts (Sat–Fri). The week of May 9–May 15, 2026 had 227 posts, and the week before had 245.
Bar chart of weekly #RStats posts (Sat–Fri). The week of May 2–May 8, 2026 had 245 posts, and the week before had 303.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Apr 25–May 1, 2026 had 303 posts, and the week before had 313.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Apr 18–Apr 24, 2026 had 313 posts, and the week before had 443.
Group of women running on a street, red arrow points to one of the women toward the back of that pack.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Apr 11–Apr 17, 2026 had 443 posts, and the week before had 331.
Bar chart titled “Figure 2. Views on Trump’s Handling of Immigration, 2025 and 2026, by Religious Affiliation.” Subtitle: “Percent who ______ handling immigration:” Blue bars are March 2025, “Approve of the job President Trump is doing.” Green bars are February 2026, “Hold favorable views of how Trump is” handling immigration. Values shown: All Americans 48 in 2025, 35 in 2026. White evangelical Protestant 78, 69. White Catholic 63, 53. White mainline/non-evangelical Protestant 64, 46. Jewish 39, 30. Hispanic Protestant 34, 23. Other non-Christian religion 43, 23. Unaffiliated 35, 21. Hispanic Catholic 29, 19. Black Protestant 29, 15. Source: PRRI Surveys, 2025 and 2026. [Alt text by ChatGPT]
Update (April 6, 2026): I used vitals and the same three-item dataset to test several LLMs from Google’s new Gemma 4 open-weight, commercially permissive family announced on April 2. While the 4b version did about the same as other local LLMs I’ve tried, gemma-4-26b scored a surprising 100% when I ran it six times. Note that although it ran at an acceptable speed in Ollama, gemma-4-26b was a tight fit in my PC’s memory when running vitals inside RStudio. In fact, it choked when I tried to run multiple epochs at once, so I ended up running only one test at a time. Also important: I set up the ellmer chat object to turn off the model’s “thinking.” The code: chat_gemma_26b <- chat_ollama( model = "gemma4:26b", api_args = list( think = FALSE ) ) my_task_gemma_26b_1 <- Task$new( dataset = my_dataset, solver = generate(chat_gemma_26b), scorer = model_graded_qa( scorer_chat = chat_anthropic(model = "claude-opus-4-6") ), name = "Basic gemma26b test running just once with Opus judging" ) my_task_gemma_26b_1$eval(epochs = 1)
Bar chart of weekly #RStats posts (Sat–Fri). The week of Mar 28–Apr 3, 2026 had 298 posts, and the week before had 224.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Mar 21–Mar 27, 2026 had 224 posts, and the week before had 341.
summarytools provides a coherent set of functions centered on data exploration and simple reporting. At its core reside the following four functions: Function Description freq() Frequency Tables featuring counts, proportions, cumulative statistics as well as missing data reporting ctable() Cross-Tabulations (joint frequencies) between pairs of discrete/categorical variables, featuring marginal sums as well as row, column or total proportions descr() Univariate (‘Descriptive’) Statistics for numerical data, with common measures of central tendency and dispersion dfSummary() Data Frame Summaries featuring type-specific information for all variables: univariate statistics and/or frequency distributions, bar charts or histograms, as well as missing data counts and proportions. Very useful to quickly, detect anomalies and identify trends at a glance
Bar chart of weekly #RStats posts (Sat–Fri). The week of Mar 14–Mar 20, 2026 had 341 posts, and the week before had 289.
Bar chart of weekly #RStats posts (Sat–Fri). The week of Mar 7–Mar 13, 2026 had 289 posts, and the week before had 276.
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 45 yesterday and 57 the day before.
New York Times AI Initiatives slide: “Our principles for using A.I. in the Newsroom.” “1. As a tool in service of our mission” “To assist our journalists in uncovering the truth and helping people understand the world.” “2. With human guidance and review” “We don’t use A.I. to write articles, and journalists are ultimately responsible for everything that we publish.” “3. Transparently and ethically” “Our core principles of journalism apply regardless of what technology is involved.”
Paper wheel shows state names on the outer ring and then inside it points to one state (in this case Massachusetts) with info about things like: Population Median age Home value compared with the U.S. as a whole From the 2024 American Community Survey one-year estimates.
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 55 yesterday and 48 the day before.
Screen shot of app with filters on left and searchable table on right
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 40 yesterday and 55 the day before.
Dashboard-style web app titled "NICAR 2026 Posts on Bluesky from Jan. 13 to Feb. 26." At the top are three large summary tiles showing total posts, unique authors, and total likes. A left sidebar lets you select a date range and filter by authors (with checkboxes) and optionally show only posts containing external URLs, plus a button to download the current dataset. The main area contains a searchable, filterable table of posts with columns for timestamp, post text snippet, author, likes, reposts, replies, and tags; entries include links to open the original post. [Alt text by ChatGPT]
Bird in a snowy bush
A bowl of vegetable bean soup with kale.
Snow covered trail in the woods, with snow covered trees on either side
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 52 yesterday and 44 the day before.
Hex-shaped logo showing a map of the continental US inside a talk bubble, with "censuschat" below
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 44 yesterday and 55 the day before.
Small portion of the app that shows Typeless, allowing toggling.
Screenshot of a NICAR 2026 Session Explorer interface showing filtered results for data visualization sessions on Friday morning. On the left panel, a user question reads: "Are there any dataviz sessions Friday morning?" The AI response lists four sessions: Customizing ggplot for yourself or your organization Fri 10:15 am, Room 209 Speaker: Kaitlyn Radde, Pew Research Center Learn how to create customized ggplot graphics packages for publication-ready visualizations. Connect with skeptical and disengaged audiences … with data! Fri 11:30 am, White River G Speakers: Eva Constantaras, Lighthouse Reports; Aaron Mendelson, The Trace; Max Siegelbaum, Documented Discover innovative ways to present and explain data to engage skeptical audiences. QGIS 2: Analyzing geographic data Fri 10:15 am, Room 204 Learn to explore, manipulate, and analyze geographic datasets to gain new insights. QGIS 1: Spatial analysis for beginners Fri 9 am, Room 204 Learn to make your own maps using QGIS, including importing and displaying geographic data. A note at the bottom says: "Please verify with the official schedule." The right panel is a session directory table with the same four sessions, showing titles, times, speakers, rooms, skill levels, and recording status. [Alt text by ChatGPT]
Snow-covered trail with trees on either side. Person with green jacket in the distance on the trail.
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 36 yesterday and 44 the day before.
Incident with Actions Update – "We continue to investigate failures impacting GitHub Actions hosted-runner jobs. We're waiting on our upstream provider to apply the identified mitigations, and we're preparing to resume job processing as safely as possible." Feb 02, 2026 – 22:10 UTC Update – "Copilot is experiencing degraded performance. We are continuing to investigate." Feb 02, 2026 – 21:27 UTC Update – "We continue to investigate failures impacting GitHub Actions hosted-runner jobs. We have identified the root cause and are working with our upstream provider to mitigate. This is also impacting GitHub features that rely on GitHub Actions (for example, Copilot Coding Agent and Dependabot)." Feb 02, 2026 – 21:13 UTC Update – "The team continues to investigate issues causing GitHub Actions jobs on hosted runners to remain queued for extended periods, with a percentage of jobs failing. We will continue to provide updates as we make progress toward mitigation." Feb 02, 2026 – 20:27 UTC Update – "Pages is experiencing degraded performance. We are continuing to investigate." Feb 02, 2026 – 19:48 UTC Update – "The team continues to investigate issues causing GitHub Actions jobs on hosted runners to remain queued for extended periods, with a percentage of jobs failing. We will continue to provide updates as we make progress toward mitigation." Feb 02, 2026 – 19:44 UTC Update – "Actions is experiencing degraded availability. We are continuing to investigate." Feb 02, 2026 – 19:43 UTC Update – "GitHub Actions hosted runners are experiencing high wait times across all labels. Self-hosted runners are not impacted." Feb 02, 2026 – 19:07 UTC Investigating – "We are investigating reports of degraded performance for Actions." Feb 02, 2026 – 19:03 UTC
Update – "We continue to investigate failures impacting GitHub Actions hosted-runner jobs. We're waiting on our upstream provider to apply the identified mitigations, and we're preparing to resume job processing as safely as possible." Feb 02, 2026 – 22:10 UTC Update – "Copilot is experiencing degraded performance. We are continuing to investigate." Feb 02, 2026 – 21:27 UTC Update – "We continue to investigate failures impacting GitHub Actions hosted-runner jobs. We have identified the root cause and are working with our upstream provider to mitigate. This is also impacting GitHub features that rely on GitHub Actions (for example, Copilot Coding Agent and Dependabot)." Feb 02, 2026 – 21:13 UTC Update – "The team continues to investigate issues causing GitHub Actions jobs on hosted runners to remain queued for extended periods, with a percentage of jobs failing. We will continue to provide updates as we make progress toward mitigation." Feb 02, 2026 – 20:27 UTC Update – "Pages is experiencing degraded performance. We are continuing to investigate." Feb 02, 2026 – 19:48 UTC Update – "The team continues to investigate issues causing GitHub Actions jobs on hosted runners to remain queued for extended periods, with a percentage of jobs failing. We will continue to provide updates as we make progress toward mitigation." Feb 02, 2026 – 19:44 UTC Update – "Actions is experiencing degraded availability. We are continuing to investigate." Feb 02, 2026 – 19:43 UTC Update – "GitHub Actions hosted runners are experiencing high wait times across all labels. Self-hosted runners are not impacted." Feb 02, 2026 – 19:07 UTC Investigating – "We are investigating reports of degraded performance for Actions." Feb 02, 2026 – 19:03 UTC
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 47 yesterday and 38 the day before.
Bar chart of daily #RStats posts. Amounts range from around 25 in September 2024 to peaking over 100 some days. Latest totals were 38 yesterday and 36 the day before.
Screenshot of a YouGov survey results table titled "YouGov Survey: ICE Policies and Practices," conducted January 9–11, 2026, among 1,129 U.S. adult citizens. The table presents responses to the question: “Do you have a favorable or an unfavorable opinion of the following?” referring to Immigration and Customs Enforcement (ICE). Data is broken down by gender, age, and race. A red arrow points to the column for White respondents, under the "Race (4 category)" heading. Key data from the White respondents column: Net (Favorable - Unfavorable): -1% Very favorable: 29% Somewhat favorable: 19% Somewhat unfavorable: 16% Very unfavorable: 37% Don’t know: 1% Unweighted N for White respondents: (718). [Alt text by ChatGPT]
Video shows a poorly formatted simple R test script being properly formatted with better spacing and indenting
Screenshot of a Git branch selection menu from a graphical Git interface. The dropdown contains the following options and branch listings: Create new branch... Create new branch from... Checkout detached... There are also a number of listed branches. Alt text by ChatGPT
Screenshot of a course module interface listing topics in a Data Science program. The modules shown are: 1. Data Science: R Basics 2. Data Science: Visualization 3. Data Science: Probability 4. Data Science: Inference and Modeling 5. Data Science: Productivity Tools 6. Data Science: Wrangling 7. Data Science: Linear Regression 8. Data Science: Building Machine Learning Models 9. Data Science: Capstone [alt text by ChatGPT]