The views of TokyoR

Yesterday, I have released the package, “slideshare”.

That package enables you to get the metadata of uploaded slides (e.g. the number of views, favorites, downloads, and so on ) through the slideshare API.

Today, I will try to visualize one of the metadata with the package.

Here's the today's work.

I have got the information of slides whose tags are “tokyor”, which is added to slides in relation to “Tokyo R user group meeting”.

So What ?

There are 65 slides tagged “TokyoR”. Seven of them have been viewed over 10,000 times.

Furthermore, the top one was viewed 30,533 times!!!

TokyoR slides have been viewed 304,521 times in total!!!

So far, about 1,500 people attended TokyoR, meaning that there are 20 times more viewers than attendees.

I hope TokyoR will keep contributing to R community.

Code

library(slideshare)
library(rCharts)
# get data
test <- Slideshare$new(apikey="XXXXXXXX",sharedsecret="XXXXXXXX")
result <- test$getSlideshowByTag("tokyor", limit=100)

# prepare for plot
result <- result[,c("Created", "NumViews", "Username", "StrippedTitle")]
colnames(result) <- c("x", "y", "Username", "Title")
result$x <- paste("#!", as.numeric(as.Date(result$x))*86400000, "!#")
result$y <- as.numeric(result$y)

# visualize
h1 <- Highcharts$new()
h1$chart(type="scatter")
for(un in unique(result$Username)){
  h1$series(data=toJSONArray2(subset(result, Username==un), json = F),
            name=un)
  }
h1$xAxis(title = list(text = "Upload date"),
         type = "datetime",
         dateTimeLabelFormats = list(month= '%Y-%b')
         )
h1$yAxis(title = list(text = "Number of views"), 
         min=0,
         label=list(format="!#'{value:,.0f}'#!")
         )
h1$legend(align = 'center', 
          verticalAlign = 'top', 
          layout = 'horizontal',
          title = list(text = "usernames"))
h1$plotOptions(scatter = list(marker = list(radius=6, symbol = 'circle')))
h1$tooltip(useHTML = T, 
           formatter = "#! function() {
           return  '<p>TITLE: ' + this.point.Title + '</p>'+
           '<p>NAME: ' + this.point.Username + '</p>'+
           '<p>DATE: ' + Highcharts.dateFormat('%Y-%m-%d', this.x) + '</p>'+
           '<p>VIEWS:' + this.y + '</p>'; } !#")
h1$title(text = "TokyoR Slides")
h1$show("iframesrc", cdn=TRUE)

TokyoR, 22 February 2014

On 22 February 2014, I attended the 36th Tokyo R user group meeting (TokyoR). It was a very good turn out with about 50 attendees at the NIFTY Corporation's office in Shinjuku Tokyo.
The session started at 14:30 with six presentations and five lightning talks, followed by drinks.

Presentations
This meeting were presented in Japanese. So, I named English titles for them.

1. Learn R in 10 minutes
The first presentation was given by Nobuaki Ooshiro from Yahoo Japan. He presented "Learn R in 10 minutes" . That was very useful introduction to R.

2. An introduction to statistical power
Takashi J. Ozaki, who is famous for his blog "Data Scientist in Ginza, Tokyo", gave an insightful talk about statistical power. For data scientist wannabies, it is important to understand the concept of statistical power. He made a supplementary about his talk in his blog.

3. Reproducible research with Rmarkdown, knitr and Pandoc
@teramonagi talked about reproducible research. With Rmarkdown, knitr and Pandoc, he presented how to make reports. It was a very useful reference about reporting with R.

4. Visualization with tSNE and SPADE
@Med_KU showed the method of dimensionality reduction and visualization with two packages, tSNE and SPADE. By the method, he visualized features of anime characters. Sorry, because I'm not good at anime, I can not say that I fully understood his presentation. But I understand his enthusiasm!!! Also, he said that he is writing an e-book for data mining.

5. EasyHTMLReport makes your reporting life happier
Yohei Sato introduced his package EasyHTMLReport. This package enables us to send HTML reporting mail from R. You don't have to sweat to make a report any more. Excellent package!!!

6. Discrete choice models with R
Hiroki Sano, who works as a data scientist at Accenture, gave a talk about discrete choice models. When you works in marketing research, you can not avoid discrete choice models. He taught the mechanism of discrete choice models and how to apply it with mlogit package.

Lighting talks
After six presentations, five lightning talks were given. In only 5 minutes, all presenters talked about impressive contents.
I just list titles below. For more details, please check out slides.

• Tuning game parameters with linear programming
• Salmon cycle analysis with econometric methods
• Visualizaiton of probability distributions
• Subspace methods with R and Proce55ing
• The introduction of dplyr package
• Using Stan for artificial data

Next Tokyo R
Next meeting is scheduled for 29 March.
Thanks to NIFTY Corporation for hosting the event.