In benchmarks with 100M observations, randomtag proves 𝟏𝟎𝐱 𝐟𝐚𝐬𝐭𝐞𝐫 than sample. 👉 If you're lazy like me (and don't care about randomization) . keep in 1/1000 👉 keeping first 1000 obs. proves 𝟒𝟎𝐱 𝐟𝐚𝐬𝐭𝐞𝐫 than sample. For 𝐡𝐲𝐩𝐞𝐫𝐬𝐩𝐞𝐞𝐝, skip loading the full dataset entirely . use "data.dta" in 1/1000

❌ Note that the command 'sample' is very slow. If you can, use package 'randomtag' . ssc install randomtag . randomtag, count(1000) gen(t) . keep if t == 1 . *This samples 1000 observations. 👉 If you want 5% sample instead . local sh = round(_N/5) . randomtag, count(`sh') gen(t) . keep if t == 1

Thanks for 4 followers! I wonder what the ratio of Bluesky like to Twitter likes is like? #newtobsky#ishouldgotosleep#4likesformetogotobed#improbablyalresdyasleep#randomtag

Felt cute, might delete later. Had a bit of an Aerith Gainsborough moment with this dress, ngl. Was planning on exchanging this for a black color, but I'm slowly vibing with this color more and more~ #clothinghasnogender#aerithVibes#ff7#sailorVenus#nerdxdress#randomTag lol