3 ms·
Congratulations, this is the most pedantic comment I've read on HN in quite a while (that's saying something). At the end of the day, log vs. linear is a visua
by rjdagost 12y ago
Congratulations, this is the most pedantic comment I've read on HN in quite a while (that's saying something). At the end of the day, log vs. linear is a visualization preference. It's not like applying a log transform to data adds any new information to a graph. Seems a bit rigid to question the author's entire piece because he doesn't use the plot scale you like.
- ars 12y agoCompare: Non-logarithmic: https://upload.wikimedia.org/wikipedia/commons/0/01/United_States_Income_Distribution_1947-2007.svg https://upload.wikimedia.org/wikipedia/commons/0/01/United_S... Apparent conclusion: Highest bracket is growing much faster than the others! Logarithmic: https://upload.wikimedia.org/wikipedia/commons/archive/0/01/20110222155802!United_States_Income_Distribution_1947-2007.svg https://upload.wikimedia.org/wikipedia/commons/archive/0/01/... Correct conclusion: All the brackets are growing identically. The scale actually matters a lot.
- Hytosys 12y agoIf you remove any 4 lines from either of your graphs, the scale doesn't matter. There is no logarithmic relation between any number of series in the linked article because there is only one series in any figure. As such, a linear scale works just fine.
- ars 12y agoHu? If you remove 4 lines from the graph the graph is blank. I'm not trying to link number series in the original article, I'm trying to decide if the dollar volume is indeed flat like he says, but I can't tell by looking at the graph.