Steven Clarke (@StevenRJClarke) profile picture
Joined March 2019
Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)Photo by Steven Clarke (@StevenRJClarke)
Replying to @DataGreg
@DataGreg I've been using your article on visualising H2O decision trees in R. Thanks it's v useful. What I am confused about is that in both datasets, you have a binary response variable; yet the tree is giving negative predictions. Why is this happening? (Airline is bernoulli)
5
Replying to @British_Airways
@British_Airways my girlfriend is travelling Edinburgh->London->Delhi but no longer needs to get on at Edinburgh. Why can’t you let her take her seat at London and save her money and hassle?
Replying to @sogoodlord
@sogoodlord My girlfriend is having to delay her move because you are indecisive over a shortfall agreement. She has removal van planned and has paid to extend her family's visa to help with move, which will now be wasted.
Replying to @dsquareddigest
@dsquareddigest I appreciate your Simply Red boosterism. They had an excellent concert in Plymouth this weekend. Great to be there with my Mum and Dad
1
Replying to @harryhutton
@harryhutton Hi Harry. Please could I be invited to your blog?
Replying to @dsquareddigest
@dsquareddigest Hi Dan. Unable to comment on the housing thread. What I don’t understand: if real wages increase rents, then house prices have to increase even more to make yields stay in line with low interest rates
1
@election_data Hi Ian. Thanks for your great account. In which constituency should I campaign in an upcoming election to do the most damage to the Tories? (preferably near Wimbledon)
Replying to @DietzVollrath
@DietzVollrath Hi. I'm a huge fan of your blog. Unfortunately, the images are broken in 'The deep roots of development - part 2': growthecon.com/blog/Deep-Roo… I would appreciate it if you could fix them. Many thanks.
1
Replying to @kjhealy
@kjhealy rel_by_region <- gss_sm %>% group_by(bigregion, religion) %>% summarize(N = n()) %>% mutate(freq = N / sum(N), pct = round((freq*100), 0)) Factor `religion` contains implicit NA, consider using `forcats::fct_explicit_na` Please help