Skip to content

FAQ.md

Latest commit

 

History

History
128 lines (83 loc) · 7.77 KB

FAQ.md

File metadata and controls

128 lines (83 loc) · 7.77 KB

Frequently Asked Questions

What Excel tutorials do you recommend?

The following resources are helpful:

The last resource offers videos on narrow topics (e.g., using the VLOOKUP function).

How do I move to the top of a selected range in Excel?

After selecting a range (for example, with Ctrl+Shift+Down), press Enter to move to the top of the range.

How do I clear the contents of a range in Excel for Mac?

Select the range, then press fn+Delete.

Can you recommend authors to follow for the topic of technology and organization design?

Each of these authors regularly publishes interesting articles or books on the topic:

How can I access Tableau?

Tableau Software provides a free license to students. Sign up via the form at the bottom of this page.

How can I get started with Python?

The Anaconda Platform provides a comprehensive solution for analyzing data with Python. The standard version of Anaconda includes common tools such as NumPy, pandas, StatsModels, and scikit-learn. In addition Jupyter Notebook, included with Anaconda, provides one of the best environments for working with data in Python.

There are many resources online for learning Python. Automate the Boring Stuff with Python is an excellent general introduction to Python. The following sites are useful for learning data analysis with Python.

For Coursera, choose the "audit" option to view course materials. Note that you need a Coursera account to select the audit option. The site offers a paid service, but you can access course materials for free by auditing.

How can I get started with R?

I recommend using RStudio and the tidyverse approach. Tidyverse is a group of packages that add useful data manipulation and visualization tools. Please see the Resources module on Carmen for directions for installing R, RStudio (an integrated development environment for R), and tidyverse.

Tidyverse includes a package, dplyr, that simplifies the syntax for most common data manipulations. As an example, suppose we have a dataset called dat that describes products a retailer sells. The dataset has two variables: price and category. To calculate the average price for each product category in base R:

with(dat, aggregate(price, list(category), mean))

The with function specifies that we use variables from the dat dataset. The aggregate function groups the dataset by the values of category and applies the function mean to each group.

Using dplyr, the code is:

dat %>%
  group_by(category) %>%
  summarize(mean(price))

The pipe operator %>% feeds the output of the expression on its left to the function on its right. We read this as "take the dat dataset, group it by product category, and produce a summary of the mean price by group." The dplyr syntax is easier to read and write than base R for most users.

An excellent resource for learning R with tidyverse is R for Data Science by Grolemund and Wickham. The book is available online for free. Wickham created the dplyr package as well as other prevalent packages, such as the ggplot2 package for data visualization.

In addition, two good courses for learning base R are:

For Coursera, choose the "audit" option if you enroll. The site pushes the 7-day trial of the paid service, but you can access all course materials for free by auditing.

My advice for learning R is to begin with R for Data Science chapters 1 through 6, the DataCamp course, or the first two weeks of the Coursera course. After the introduction, you will be able to implement MBA 6273 exercises using R. It's hard to learn a programming language in isolation, and a moving target (i.e., exercises of increasing difficulty) helps. I'm happy to discuss any specific questions you have while learning R.

Should I Learn SQL?

I recommend starting with Python or R. The Blaze package for Python and the dplyr package for R implement core database operations using terms reminiscent of SQL. Once you are comfortable with data manipulation in Python or R, it is straightforward to translate your ideas to SQL.

For example, the dplyr database backend, dbplyr, communicates directly with standard data stores so you can take advantage of the speed and storage of a database. Specifically, dbplyr translates your R code to the appropriate database query. The interface supports all common databases (MySQL, SQLite, PostgreSQL, Oracle, general ODBC) and distributed storage and processing platforms like Hadoop and Spark.

Suppose we run a service business. We have a database with two data tables: sales and staff. Sales contains four variables: location, date, hour, and revenue. Staff includes three variables: location, date, and headcount.

We want to calculate labor productivity as daily revenue divided by headcount for each location and day. We must:

  1. Group revenue observations in the sales table by location and date, then total revenue within each group. E.g., for the group defined by location A and date 1/1/2020, daily revenue is the sum of all hourly revenues within the group

  2. Merge the two tables according to location and date

  3. Calculate labor productivity.

In R:

sales %>% 
  group_by(location, date) %>%
  summarize(dailyRevenue = sum(revenue)) %>%
  left_join(staff) %>%
  mutate(laborProductivity = dailyRevenue / headcount)

In SQL:

SELECT *,
       dailyRevenue / headcount AS laborProductivity
FROM staff
  NATURAL LEFT JOIN (
    SELECT location, date, hour,
           SUM(revenue) AS dailyRevenue
    FROM sales
    GROUP BY location, date
  );

In both cases, the grouping (group by) and merging (left join) verbs are similar. The R code is easier to read since it follows the logic of the operations. We select the sales table, group the sales table by location and date, summarize each group by taking the sum of revenue, merge the result with the staff table, and calculate labor productivity.