recentpopularlog in
« earlier  
The Internet of Beefs
Fascinating essay on the nature of culture war specifically as it plays out in online outrage mob culture.

"The semantic structure of the Internet of Beefs is shaped by high-profile beefs between charismatic celebrity knights loosely affiliated with various citadel-like strongholds peopled by opt-in armies of mooks."
culture  conflict  internet  history  Philosophy 
9 weeks ago
Macbook Air mid-2011 Which OS? | MacRumors Forums
Good discussion on the merits of which OS to run an old mid-2011 Macbook Air like mine. Seems like the consensus is split between 10.9 Mavericks (for speed and stability) and 10.11 El Capitan (for most modern feature set possible without bogging the old machine down).
macbookair  mac  osx  computers 
10 weeks ago
.050 vs .058 in Chainsaws
In a nutshell, .050 came from German and American manufacturers, while .058 came from Swedish manufacturers. Today .050 is more prevalent with more chain options overall. No real reason to switch other than consistency or chain choice. No real difference in performance.
chainsaw  wood  forestry 
december 2019
(92) Logistic Regression in R, Clearly Explained!!!! - YouTube
This was the key to my paper. This tutorial taught me how to use R to perform logistic regression. I also used the associated script, linked in the description. This provided the fundamental learning for the statistical analysis I did for my final Harvard paper for the Intelligence and International Security class, fall 2019.
statistics  rlanguage  rstudio  math  tutorial  howto  logisticregression 
december 2019
logistic_regression_demo/logistic_regression_demo.R at master · StatQuest/logistic_regression_demo · GitHub
This was the key to my paper. I used this script (and the associated tutorial on YouTube) as the basis for how I wrote the script for my final Harvard paper for the Intelligence and International Security class, fall 2019.
rlanguage  rstudio  statistics  math  programming  logisticregression 
december 2019
Imputing Missing Data with R; MICE package | DataScience+
This wasn't the method I chose. The method I was using was the one linked to in this article under the Update section. However, it's worth saving this approach in case I need it another day.
rlanguage  rstudio  statistics  MICEpackage 
december 2019
Handling missing data with MICE package; a simple approach | DataScience+
I attempted to use this how to almost exactly to a T to impute the missing data in my dataset for my Harvard Intelligence & International Security final. Unfortunately I kept running into errors. The final error that put a stop to this attempt basically meant my data was "not invertable" i.e. too strongly linearly correlated to impute values. So i ended up abandoning this method and simply taking the mean. Which reduces variance, unfortunately, but it was the only option left other than leaving out a bunch of countries I didn't want to omit.
statistics  rlanguage  rstudio  MICEpackage 
december 2019
Dealing with Missing Values · UC Business Analytics R Programming Guide
I used this code example to do a quick and dirty imputation with mean values. Had to do that because using MICE to do random forest imputation was not possible with my dataset. Too strongly linearly correlated. So this had to do.

"If we want to recode missing values in a single data frame variable we can subset for the missing value in that specific variable of interest and then assign it the replacement value. For example, here we recode the missing value in col4 with the mean value of col4."

df$col4[is.na(df$col4)] <- mean(df$col4, na.rm = TRUE)
df
rlanguage  rstudio  statistics  math  programming 
december 2019
« earlier      
per page:    204080120160

Copy this bookmark:





to read