Weitere ähnliche Inhalte Ähnlich wie Aargh! I have to teach R (Experiences in the teaching of R) (20) Kürzlich hochgeladen (20) Aargh! I have to teach R (Experiences in the teaching of R)1. © 2010 – 2016 eoda GmbHMartin Schneider
I have to teach R
Experiences in the teaching of R
eRum 2016
Martin Schneider
Data Scientist
2. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
About eoda
Interdisciplinary Team
Statisticians | Engineers | Economists | Sociologist | …
Based in Kassel - Germany
Data Science Consulting, Training, Support,
Software and Analytic Services with a focus on R
3. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Teaching R since
2013
Motivation and Qualification
Number of courses
0025
Course Participants
0125I
4. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Motivation and Qualification
Teaching R since
2010
Number of courses
0150
Course Participants
1000WE
5. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
BusinessAcademics
Wide differing Background
6. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
Business
Wide differing Priorities
We want to do everything perfect, every
prerequisite for analysis has to be met.
Academics
7. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
BusinessAcademics
Wide differing Priorities
How much money can we save?
We want to see results
8. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
? Excel
Wide differing preknowledge in Analytic Tools
9. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
Wide differing Statistical Background
NONECOOL
NONE
10. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
Wide differing Statistical Background
A LOT
COOL
A LOT
11. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
Wide differing Statistical Background
SOME
Dangerous
half knowledge
SOME
12. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
Wide differing Personal Goals
I want to start with
statistical analysis
I‘m already using SPSS
but I heard a lot of R so I
wanted to take a look
13. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
For who?
BusinessAcademics
Wide differing Personal Goals?
I want to start with
statistical analysis
I‘m already using SPSS
but I heard a lot of R so I
wanted to take a look
I‘m already using Excel and
everything is working fine
but my boss said…
14. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Package updates
Recurring Problems
15. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Recurring Problems
diverse classes
16. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Recurring Problems
In R there are usually lots of different ways
to solve a problem
We did this already this way, why
are we doing it another way?
17. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Recurring Problems
Steep learning curve
What does this Error Message
mean? I‘m having problems
solving this exercise.
This is in SPSS much easier
I will never
under-
stand R
18. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Recurring Problems
Technical hurdles
I´m missing the permission
to install packages
I have to use external
software
19. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Writing code
in front
Excersises
Usual Course
Small Part of Presentation
Presentation
20. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Usual Course
Writing code in front
cars <- read.csv(file = "Data/carsMO.csv", stringsAsFactors = FALSE)
select(cars, Manufacturer, Modell, contains("sales", ignore.case = TRUE))
select(cars, Manufacturer, Modell, Sales:Sales_grouped)
mutate(cars, weight_to_hp = HP/Weight) %>%
filter(weight_to_hp > mean(weight_to_hp, na.rm = TRUE)) %>%
arrange(desc(weight_to_hp))%>%
select(Modell, weight_to_hp)
abs_loss_data <- cars %>%
mutate(absolute_loss = Purchase_price - Resaleprice) %>%
arrange(desc(absolute_loss)) %>%
select(Modell, contains("price"), absolute_loss) %>%
filter(!is.na(absolute_loss))
Slowdown
when writing
code
21. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Usual Course
Excercises
Don‘t make it
to difficult
Now that you know, what a list is
please load this dataset, apply a
cluster analysis on it and create
an appealing report
22. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Usual Course
Excercises
Don‘t make it
to easy
Now that you know how to
extract the first element from a
list, please extract the second
23. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Example Data
Iris
NYC Taxi Data
MT Cars
Usual Course
24. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
Wishfull Thinking
workshop element attached to the course
smaller chunks of training sessions
25. © 2010 – 2016 eoda GmbHMartin Schneider www.eoda.de
@eodaGmbH
@eodaGmbH eodaGmbH
blog.eoda.de
eoda GmbH
Universitätsplatz 12
34127 Kassel - Germany
www.eoda.de/en
info@eoda.de
+49 561 202724-40
The Data Science Specialists.