Tidy Tuesday live screencast: Analyzing IKEA furniture in R

preview_player
Показать описание
I'll analyze a dataset about the IKEA furniture, without looking at the dataset in advance.

Рекомендации по теме
Комментарии
Автор

1:34 Downloading and exploring the dataset [ikea <- %>% select(-X1)]
4:06 What are the most common categories? [count(), fct_reorder(category, n), geom_col(), labs()]
5:12 Distribution of prices per category. [geom_boxplot(), scale_x_log10(labels = dollar), glue("{ category } (n})"]
9:05 Converting Saudi Riyal to USD.
11:01 Visualize same data with Joy Plot to show bimodality. [library(ggridges), geom_density_ridges()]
15:30 Use [fill = other_colors] to investigate whether extra colors increase the price.
17:03 Building a predictive model of category, other_colors, volume.
17:46 Are there items that appear in multiple categories. [count(name, category, sort = TRUE)]
19:20 Visualize results. [geom_col(), fct_lump(), fct_reorder(name, n, sum),
23:00 Using short_description information. [str_trim(), str_replace_all(), separate(), extract()]
29:48 Unite() category and main_description.
31:47 Calculate the volume in cubic cm/liters/cubic meters.
34:23 What are the biggest objects? [geom_boxplot()]
39:25 What is the highest price per cubic meter? [color = category, fct_lump(), geom_point(), geom_smooth()]
44:13 str_squish() instead of str_trim() is very useful.
45:08 Investigating how price relates to designer. [group_by(), summarize(), n_distinct()]
46:46 Creating a model for volume, category, price. [lm(), fct_relevel()]
52:50 Coefficient/Tie fighter plot. [library(broom), tidy(), ggplot(aes(estimate, term)), geom_errorbarh()]

TheDataDigest
Автор

This is actually the greatest DS content out there on YouTube. Thank you, David!

LazySnake
Автор

Dude this is amazing. Thank you so much for explanations!

Rumil_
Автор

Thanks for the video, good work, learn a lot

taiwankyh
Автор

These are great! Thanks, David! For the community - any chance there is a similar feed for those more comfortable in Python?

kgiff