danniemcq
danniemcq.bsky.social
danniemcq
@danniemcq.bsky.social
I do data analysis for fun and occasionally work too
Run it again, it might work this time
December 1, 2024 at 2:41 PM
Found the owners eventually but our goats have guests in the field overnight!
February 18, 2024 at 6:06 PM
Even if you're on life support?
February 16, 2024 at 12:05 AM
Trying to make people think they are a big writer and not *checks notes* the cuts financial advice columnist
February 16, 2024 at 12:04 AM
Answer Spoiler if you are doing the course so be warned

N <- 10000
runCount <- 0
for(i in rnorm(N)){
if (i >= -1 & i <= 1){
runCount <- runCount + 1
}
}
ratio <- runCount/N
ratio
February 15, 2024 at 9:34 PM
if i do it 1000 times i'd expect it to get to about 50/50. 1000000 times even more so.

with that using rnorm its expected to get a number between -1 and 1 68.2% of time

n=10 got .3
n=1000 got .693
n=100000 got .68099
n=10000000 got .6828726
n=1000000000 got .682669
February 15, 2024 at 9:29 PM
First big thing is the law of large numbers, the more times you do something the more likely you can predict the results. So IF something is outta whack of the ordinary you can spot it. Like flipping a coin, if i do it twice it could be 2 heads or 100% but thats somewhat normal but...
February 15, 2024 at 9:13 PM