Count the usefulness by learning platform.
# R code provided for user
library(tidyverse)
df <- read_csv("multipleChoiceResponses1.csv")
use_cols <- df %>% select(starts_with("LearningPlatformUsefulness"))
long <- use_cols %>%
pivot_longer(everything(), names_to="learning_platform", values_to="usefulness") %>%
drop_na() %>%
mutate(learning_platform = str_replace(learning_platform, "LearningPlatformUsefulness",""))
table1 <- long %>% count(learning_platform, usefulness)
table1Result:
learning_platform usefulness n
Arxiv Not Useful 37
Arxiv Somewhat useful 1038
Arxiv Very useful 1316
Blogs Not Useful 45
Blogs Somewhat useful 2406
Blogs Very useful 2314
College Not Useful 101
College Somewhat useful 1405
College Very useful 1853
Communities Not Useful 16
Communities Somewhat useful 567
Communities Very useful 559
Company Not Useful 41
Company Somewhat useful 502
Company Very useful 438
Conferences Not Useful 119
Conferences Somewhat useful 1305
Conferences Very useful 758
Courses Not Useful 47
Courses Somewhat useful 1750
Courses Very useful 4195
Documentation Not Useful 42
Documentation Somewhat useful 1067
Documentation Very useful 1212
Friends Not Useful 51
Friends Somewhat useful 774
Friends Very useful 756
Kaggle Not Useful 56
Kaggle Somewhat useful 2451
Kaggle Very useful 4076
Newsletters Not Useful 56
Newsletters Somewhat useful 686
Newsletters Very useful 347
Podcasts Not Useful 124
Podcasts Somewhat useful 818
Podcasts Very useful 272
Projects Not Useful 39
Projects Somewhat useful 1185
Projects Very useful 3570
SO Not Useful 64
SO Somewhat useful 2044
SO Very useful 3532
Textbook Not Useful 69
Textbook Somewhat useful 1796
Textbook Very useful 2316
TradeBook Not Useful 9
TradeBook Somewhat useful 162
TradeBook Very useful 162
Tutoring Not Useful 32
Tutoring Somewhat useful 569
Tutoring Very useful 825
YouTube Not Useful 104
YouTube Somewhat useful 2605
YouTube Very useful 2520tot <- long %>% count(learning_platform, name="tot")
use <- long %>% filter(usefulness!="Not Useful") %>%
count(learning_platform, name="count")
merged <- left_join(use, tot)
merged <- merged %>% mutate(perc_usefulness = count/tot)
mergedResult:
learning_platform count tot perc_usefulness
Arxiv 2354 2391 0.984525
Blogs 4720 4765 0.990556
College 3258 3359 0.969932
Communities 1126 1142 0.985989
Company 940 981 0.958206
Conferences 2063 2182 0.945463
Courses 5945 5992 0.992156
Documentation 2279 2321 0.981904
Friends 1530 1581 0.967742
Kaggle 6527 6583 0.991493
Newsletters 1033 1089 0.948577
Podcasts 1090 1214 0.897858
Projects 4755 4794 0.991865
SO 5576 5640 0.988652
Textbook 4112 4181 0.983497
TradeBook 324 333 0.972973
Tutoring 1394 1426 0.977560
YouTube 5125 5229 0.980111top10 <- merged %>% arrange(desc(count)) %>% head(10)
# compute cumulative pct
total_count <- sum(merged$count)
top10 <- top10 %>% mutate(cum_pct = cumsum(count)/total_count)
other <- sum(merged$count) - sum(top10$count)
table3 <- bind_rows(top10, tibble(learning_platform="Other",count=other,cum_pct=1))
table3Result:
learning_platform count tot perc_usefulness cum_pct
Kaggle 6527 6583.0 0.991493 0.120533
Courses 5945 5992.0 0.992156 0.230319
SO 5576 5640.0 0.988652 0.333290
YouTube 5125 5229.0 0.980111 0.427933
Projects 4755 4794.0 0.991865 0.515743
Blogs 4720 4765.0 0.990556 0.602907
Textbook 4112 4181.0 0.983497 0.678842
College 3258 3359.0 0.969932 0.739008
Arxiv 2354 2391.0 0.984525 0.782479
Documentation 2279 2321.0 0.981904 0.824565
Other 9500 NaN NaN 1.000000Plotting based on results.
library(ggplot2)
ggplot(top10, aes(x=fct_reorder(learning_platform, count), y=count)) +
geom_col() +
coord_flip() +
labs(title="Top 10 Useful Learning Platforms", x="Platform", y="Count")