R02_Prac01_dplyr_ptt

1 Loading pkgs and options

# add essential pkgs here
library(tidyverse)

## ── Attaching packages ──────────────────────────────────── tidyverse 1.2.1 ──

## ✔ ggplot2 3.1.0     ✔ purrr   0.2.5
## ✔ tibble  1.4.2     ✔ dplyr   0.7.8
## ✔ tidyr   0.8.2     ✔ stringr 1.3.1
## ✔ readr   1.1.1     ✔ forcats 0.3.0

## ── Conflicts ─────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

options(stringsAsFactors = F)

2 Loading rda data

• will creating two data frames, posts and comments.

load("data/HatePolitics_Hang_20190101_2019_06_01.rda")

3 Counting the top 5 commentors

• 找出回文數最高的五個人

    # you cade should be here
    slice(1:5)

## # A tibble: 5 x 2
##   commentor        n
##   <chr>        <int>
## 1 gn02118620    7362
## 2 zeumax        5908
## 3 ninaman       5237
## 4 bottger       5062
## 5 bluesunflowe  4421

4 Counting the top 5 posters

• 找出發文數最高的五個人

    # you cade should be here
    # you cade should be here
    slice(1:5)

## # A tibble: 5 x 2
##   poster        n
##   <chr>     <int>
## 1 mark2165    198
## 2 kolod546    106
## 3 sincsnow     86
## 4 c1951        85
## 5 fantasy14    82

5 Enumerating comment order for each posts

• 產生兩個新的變項，cindex指出該則回文為該篇貼文的第幾則回文，ctotal指出該則回文所在的貼文共有幾則回文。
• enumerating the order, the “floor”, of each comment
• Who are always reponse to Hang’s posts very soon?
• Practice goal: Using group_by and mutate()
• comment order should be cindex
• total comments of a post should be ctotal
• You must use row_number() to get the order of each comment.

    # you cade should be here
    # you cade should be here
    select(tag, commentor, cindex, ctotal) %>%
    head(n=20)

## # A tibble: 20 x 4
##    tag   commentor    cindex ctotal
##    <chr> <chr>         <int>  <int>
##  1 "→ "  tarcowang         1    103
##  2 "推 " alex8725          2    103
##  3 "→ "  lovebxcx          3    103
##  4 "推 " want150           4    103
##  5 "→ "  want150           5    103
##  6 "→ "  lovebxcx          6    103
##  7 "→ "  dai26             7    103
##  8 "→ "  lovebxcx          8    103
##  9 "→ "  lovebxcx          9    103
## 10 "→ "  lovebxcx         10    103
## 11 "推 " f124             11    103
## 12 "推 " willy0526        12    103
## 13 "→ "  lovebxcx         13    103
## 14 "→ "  dodolong0310     14    103
## 15 "噓 " poolplayer       15    103
## 16 "推 " Stigmata         16    103
## 17 "推 " YINCHAUN         17    103
## 18 "噓 " simon78410       18    103
## 19 "→ "  lovebxcx         19    103
## 20 "→ "  lovebxcx         20    103

6 Calculating the average response order for each commentor

• 延續上題，計算出每一個人回應的平均樓層（avg_order 比方說，某人共在三篇貼文的回應樓層為第3樓、第6、第9樓，其平均回應樓層為6）；如果除以該則貼文的總回文樹的話，就會變成平均回文樓層比例（以下範例為有除以總回文數的結果）。另外順便計算標準差（sd_order）以及每個人的總回文數量tot_comment。
• Filter out those posts with fewer than 5 comments 不列計少於5則回文的貼文
• Filter out those commentors with fewer than 10 comments 不列計在資料中回文總數少於10則的回文者
• Only show top 10 commentors who always respond very fast

    # you cade should be here
    # you cade should be here
    slice(1:10)

## # A tibble: 10 x 4
##    commentor    avg_order sd_order tot_comment
##    <chr>            <dbl>    <dbl>       <int>
##  1 jma306          0.0549   0.0753          13
##  2 ckbdfrst        0.0898   0.0537          14
##  3 popy8789        0.0961   0.0665          15
##  4 flavorBZ        0.0969   0.0833          18
##  5 wangyc          0.105    0.0621          14
##  6 GalLe5566       0.127    0.0566          14
##  7 WeiYinChen16    0.133    0.176           16
##  8 hachime         0.142    0.0998          20
##  9 Rrrxddd         0.156    0.202           32
## 10 jk952840        0.159    0.100           14

7 Challenge: Counting how many posts each commentor has responded to

針對每一個回文者，計算出下列數值 1. 一共回過多少則post，一個post多個回文仍只算一次 2. 針對每篇post，最多曾回應一篇post幾次 3. 針對每篇post，最少曾回應一篇post幾次 4. 針對每篇post，平均回應過幾次 5. 針對每篇post，回應次數的標準差

    # you cade should be here
    # you cade should be here
    slice(1:30)

## # A tibble: 30 x 6
##    commentor    num_post max_c2p min_c2p mean_c2p sd_c2p
##    <chr>           <int>   <dbl>   <dbl>    <dbl>  <dbl>
##  1 aylao            1775      29       1     4.27   5.45
##  2 bottger          1569     148       1    15.0   27.2
##  3 ninaman          1497      87       1    16.3   18.1
##  4 zeumax           1286     110       1    17.4   22.1
##  5 gn02118620       1206     128       1    36.4   36.7
##  6 bluesunflowe     1197      86       1    18.3   21.3
##  7 pupu20317        1190     103       1    15.6   24.6
##  8 formatted        1114      77       1    15.6   18.6
##  9 alex8725          984      64       1    12.1   13.9
## 10 c1951             886      38       1     6.19   7.05
## # ... with 20 more rows

8 Call for idea

• Except the above methods, try to propose your idea to show questionable users, possible big fans, and high frequency enthusiasts.

# your idea here
# your code here