|
| 1 | +### Please add alt text to your posts |
| 2 | + |
| 3 | +Please add alt text (alternative text) to all of your posted graphics for `#TidyTuesday`. |
| 4 | + |
| 5 | +Twitter provides [guidelines](https://help.twitter.com/en/using-twitter/picture-descriptions) for how to add alt text to your images. |
| 6 | + |
| 7 | +The DataViz Society/Nightingale by way of Amy Cesal has an [article](https://medium.com/nightingale/writing-alt-text-for-data-visualization-2a218ef43f81) on writing _good_ alt text for plots/graphs. |
| 8 | + |
| 9 | +> Here’s a simple formula for writing alt text for data visualization: |
| 10 | +> ### Chart type |
| 11 | +> It’s helpful for people with partial sight to know what chart type it is and gives context for understanding the rest of the visual. |
| 12 | +> Example: Line graph |
| 13 | +> ### Type of data |
| 14 | +> What data is included in the chart? The x and y axis labels may help you figure this out. |
| 15 | +> Example: number of bananas sold per day in the last year |
| 16 | +> ### Reason for including the chart |
| 17 | +> Think about why you’re including this visual. What does it show that’s meaningful. There should be a point to every visual and you should tell people what to look for. |
| 18 | +> Example: the winter months have more banana sales |
| 19 | +> ### Link to data or source |
| 20 | +> Don’t include this in your alt text, but it should be included somewhere in the surrounding text. People should be able to click on a link to view the source data or dig further into the visual. This provides transparency about your source and lets people explore the data. |
| 21 | +> Example: Data from the USDA |
| 22 | +
|
| 23 | +Penn State has an [article](https://accessibility.psu.edu/images/charts/) on writing alt text descriptions for charts and tables. |
| 24 | + |
| 25 | +> Charts, graphs and maps use visuals to convey complex images to users. But since they are images, these media provide serious accessibility issues to colorblind users and users of screen readers. See the [examples on this page](https://accessibility.psu.edu/images/charts/) for details on how to make charts more accessible. |
| 26 | +
|
| 27 | +The `{rtweet}` package includes the [ability to post tweets](https://docs.ropensci.org/rtweet/reference/post_tweet.html) with alt text programatically. |
| 28 | + |
| 29 | +Need a **reminder**? There are [extensions](https://chrome.google.com/webstore/detail/twitter-required-alt-text/fpjlpckbikddocimpfcgaldjghimjiik/related) that force you to remember to add Alt Text to Tweets with media. |
| 30 | + |
| 31 | +# CHIP Dataset |
| 32 | + |
| 33 | +The data this week comes from [CHIP Dataset](https://chip-dataset.vercel.app/). |
| 34 | + |
| 35 | +- [Moore's Law](https://www.synopsys.com/glossary/what-is-moores-law.html#:~:text=Definition,as%20E%20%3D%20mc2).) |
| 36 | +- [Wikipedia - Moore's Law](https://en.wikipedia.org/wiki/Moore%27s_law) |
| 37 | + |
| 38 | +> Moore's law is the observation that the number of transistors in a dense integrated circuit (IC) doubles about every two years. Moore's law is an observation and projection of a historical trend. Rather than a law of physics, it is an empirical relationship linked to gains from experience in production. |
| 39 | +
|
| 40 | +Paper for citation: [Summarizing CPU and GPU Design Trends with Product Data](https://arxiv.org/abs/1911.11313) |
| 41 | + |
| 42 | +Note that the authors prohibit resharing the dataset, so I've created a simple summary. You can easily download the full dataset at the bottom of: <https://chip-dataset.vercel.app/> |
| 43 | + |
| 44 | +> Here are some interesting findings: |
| 45 | +> |
| 46 | +> - Moore's Law still holds, especially in GPUs. |
| 47 | +> - Dannard Scaling is still valid in general. |
| 48 | +> - CPUs have higher frequencies, but GPUs are catching up. |
| 49 | +> - GPU performance doubles every 1.5 years. |
| 50 | +> - GPU performance improvement is a joint effect of smaller transistors, larger die size, and higher frequency. |
| 51 | +> - High-end GPUs tends to first use new semiconductor technologies. Low-end GPUs may use old technologies for a few years. |
| 52 | +
|
| 53 | +### Get the data here |
| 54 | + |
| 55 | +```{r} |
| 56 | +# Get the Data |
| 57 | +
|
| 58 | +# Read in with tidytuesdayR package |
| 59 | +# Install from CRAN via: install.packages("tidytuesdayR") |
| 60 | +# This loads the readme and all the datasets for the week of interest |
| 61 | +
|
| 62 | +# Either ISO-8601 date or year/week works! |
| 63 | +
|
| 64 | +tuesdata <- tidytuesdayR::tt_load('2022-08-23') |
| 65 | +tuesdata <- tidytuesdayR::tt_load(2022, week = 34) |
| 66 | +
|
| 67 | +chips <- tuesdata$chips |
| 68 | +
|
| 69 | +# Or read in the data manually |
| 70 | +
|
| 71 | +chips <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-08-23/chips.csv') |
| 72 | +
|
| 73 | +``` |
| 74 | +### Data Dictionary |
| 75 | + |
| 76 | +# `chips.csv` |
| 77 | + |
| 78 | +|variable |class |description | |
| 79 | +|:------------------------|:---------|:-----------| |
| 80 | +|date |double |Date of release | |
| 81 | +|type |character |Type of chip | |
| 82 | +|foundry |character | Creator | |
| 83 | +|vendor |character | Vendor | |
| 84 | +|process_size_nm_mean |double | Process size in nanometer | |
| 85 | +|process_size_nm_sd |double | Process size in nanometer | |
| 86 | +|tdp_w_mean |double | Thermal design profile | |
| 87 | +|tdp_w_sd |double |Thermal design profile | |
| 88 | +|die_size_mm_2_mean |double | Die size in millimeters^2 | |
| 89 | +|die_size_mm_2_sd |double |Die size in millimeters^2 | |
| 90 | +|transistors_million_mean |double | Transitor count in millions | |
| 91 | +|transistors_million_sd |double |Transitor count in millions | |
| 92 | +|freq_m_hz_mean |double | Frequency (Mhz) | |
| 93 | +|freq_m_hz_sd |double |Frequency (Mhz) | |
| 94 | +|n |integer | Total number of observations for date, type, foundry, vendor grouping | |
| 95 | + |
| 96 | +### Cleaning Script |
| 97 | + |
0 commit comments