R for Excel Users: Pivot Tables, VLOOKUPs in R

February 25, 2020
By

[This article was first published on business-science.io, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

New business and financial analysts are finding R every day. Most of these new userRs (R users) are coming from a non-programming background. They have ample domain experience in functions like finance, marketing, and business, but their tool of choice is Excel (or more recently Tableau & PowerBI).

Learning R can be a major hurdle. You need to learn data structures, algorithms, data science, machine learning, web applications with Shiny and more to be able to accomplish a basic dashboard. This is a BIG ASK for non-coders. This is the problem I aim to begin solving with the upcoming release of tidyquant v1.0.0. Read the updated ‘R for Excel Users’ on Business Science.

Let’s demo several of the new Excel features!

Excel in R: Introducing tidyquant
New Features in v1.0.0.9000 (Beta on GitHub)

In tidyquant version 1.0.0.9000 (still in beta-mode on GitHub Here, but expected CRAN release of tidyquant v1.0.0 is March 2020), I have added the following features to support new useRs transitioning from an Excel background.

New Features
To make the transition to R easier for Excel Users

Pivot Tables in R

VLOOKUP in R

Summarizing “IFS” Functions

100 + New Excel Functions

NEW Tidyverse Functions – Summarize By Time

📅 NEW API Integrations (Implementation scheduled for March)

I’ll showcase a small portion of the new features in this post. Attend Learning Lab 30 (Register Here for Free) for a real business example where I showcase Shiny and tidyquant together. You’ll see how the new tidyquant features streamline development of Financial and Business Shiny Apps.

Replication Requirements

Please use tidyquant (>= 1.0.0). Installation is recommended via GitHub until the official CRAN release (expected in March 2020).

devtools::install_github("business-science/tidyquant")

Load the following libraries.

library(tidyverse)
library(tidyquant)
library(knitr)

✅ Pivot Tables

The Pivot Table is one of Excel’s most powerful features. Honestly, when I came to R, one of the biggest things I lost was the Pivot Table – A tool used for quickly summarizing data into a digestable table. It’s now in R with pivot_table().

Excel Pivot Table

For those that may have never used the Excel Pivot Table before, the process goes something like this.

  1. Start with a raw table in “long” format where each row is a unique observation
  2. Use the Excel Pivot Table dialog to position fields into Columns, Rows, and Summarized Values
  3. The Pivot Table is returned with the data summarized into columns and rows

Pivot Tables

Excel Pivot Table is now in R

R Pivot Table

Excel’s Pivot Table now in R with pivot_table(). Let’s try it out.

First, let’s get some raw data. The FANG data set ships with tidyquant and represents the ouput of the tq_get(c("FB", "AMZN", "NFLX", "GOOG")) from 2013 to 2016. This is raw OHLCV data with adjusted stock prices downloaded from Yahoo Finance.

FANG
## # A tibble: 4,032 x 8
##    symbol date        open  high   low close    volume adjusted
##                       
##  1 FB     2013-01-02  27.4  28.2  27.4  28    69846400     28  
##  2 FB     2013-01-03  27.9  28.5  27.6  27.8  63140600     27.8
##  3 FB     2013-01-04  28.0  28.9  27.8  28.8  72715400     28.8
##  4 FB     2013-01-07  28.7  29.8  28.6  29.4  83781800     29.4
##  5 FB     2013-01-08  29.5  29.6  28.9  29.1  45871300     29.1
##  6 FB     2013-01-09  29.7  30.6  29.5  30.6 104787700     30.6
##  7 FB     2013-01-10  30.6  31.5  30.3  31.3  95316400     31.3
##  8 FB     2013-01-11  31.3  32.0  31.1  31.7  89598000     31.7
##  9 FB     2013-01-14  32.1  32.2  30.6  31.0  98892800     31.0
## 10 FB     2013-01-15  30.6  31.7  29.9  30.1 173242600     30.1
## # … with 4,022 more rows

We can summarize this information with a Pivot Table using pivot_table(.rows, .columns, .values). First, I’ll take a look to see if there are any missing (NA) values. The only trick is to use a ~ in front of any calculations. All zeros is good.

FANG %>%
    pivot_table(
        .columns = symbol,
        .values  = ~ SUM(is.na(adjusted))
    ) %>%
    kable()
AMZN FB GOOG NFLX
0 0 0 0

Next, I’ll do some financial summarizations. I’d like to take a look at percentage returns by year and quarter. This is easy to do by using stacked functions with the c() operator on .columns and .rows.

FANG %>%
    pivot_table(
        .rows    = c(symbol, ~ QUARTER(date)),
        .columns = ~ YEAR(date),
        .values  = ~ (LAST(adjusted) - FIRST(adjusted)) / FIRST(adjusted)
    ) %>%
    kable()
symbol QUARTER(date) 2013 2014 2015 2016
AMZN 1 0.0356768 -0.1547856 0.2060807 -0.0680544
AMZN 2 0.0614656 -0.0530919 0.1723923 0.1956892
AMZN 3 0.1082595 -0.0299348 0.1703285 0.1538281
AMZN 4 0.2425300 -0.0223965 0.2979913 -0.1038196
FB 1 -0.0864286 0.1010785 0.0480561 0.1162199
FB 2 -0.0254603 0.0745768 0.0502020 -0.0153369
FB 3 1.0245869 0.1613283 0.0344034 0.1233033
FB 4 0.0838954 0.0192031 0.1507422 -0.1065466
GOOG 1 0.0980851 0.0017401 0.0441873 0.0041923
GOOG 2 0.0988279 0.0143170 -0.0406450 -0.0770892
GOOG 3 -0.0134815 -0.0091132 0.1659128 0.1116688
GOOG 4 0.2634837 -0.0736798 0.2414403 -0.0009578
NFLX 1 1.0571677 -0.0297393 0.1941594 -0.0702983
NFLX 2 0.1571014 0.2081494 0.5901917 -0.1345316
NFLX 3 0.3786783 -0.0463327 0.1027844 0.0194477
NFLX 4 0.1341568 -0.2214904 0.0792602 0.2062750

A few points:

  1. Collapsing – I just used functions to collapse the daily date by YEAR() and QUARTER(). This essentially creates a new grouping variable that is a part of the date.
  2. Stacking – I stacked multiple grouping operations with the c() operator: .rows = c(symbol, ~ QUARTER(date))
  3. Summarization Details – I added multiple function calls to get the Percentage Change in the .values summarization operation. This is allowed as long as the result returns a single value.
  4. Tilde Required – For calculations (e.g. ~ YEAR(date)), I used the tilde (~) each time, which is required.
  5. Tilde Not Required – For bare column names with no calculation, a tilde (~) is not required.

My Favorite Part of Pivot Tables in R

We can easily switch Pivot Tables to provide different levels of summarization. Now I’ll quickly change to returns by year. Notice I’m using a new summarization function, PCT_CHANGE_FIRSTLAST() to save me some typing.

FANG %>%
    pivot_table(
        .rows    = symbol,
        .columns = ~ YEAR(date),
        .values  = ~ PCT_CHANGE_FIRSTLAST(adjusted)
    ) %>% 
    kable()
symbol 2013 2014 2015 2016
AMZN 0.5498426 -0.2201673 1.1907495 0.1772084
FB 0.9517858 0.4260647 0.3340983 0.1255136
GOOG 0.5495473 -0.0532416 0.4460024 0.0404130
NFLX 3.0014129 -0.0584587 1.2945491 0.1258640

✅ VLOOKUP

When I first started learning R, I couldn’t grasp how to merge / join data. It was very frustrating because I was used to Excel’s VLOOKUP function that pulled a value or a column of values as needed.

Excel VLOOKUP

VLOOKUP is an Excel function to lookup and retrieve data from a specific column in table. Here’s how the process works in Excel.

  1. Start with a Lookup Table. Contains Key-Value pairs.
  2. Simple Case – Use a VLOOKUP to input a single value and output a single value.
  3. More Powerful Case – Use a VLOOKUP to add a column to an Excel Table.

VLOOKUP

Excel VLOOKUP() is now in R

R VLOOKUP

The most popular Excel Reference Function, VLOOKUP, is now in R as VLOOKUP(). It’s vectorized, which means we can use VLOOKUP() inside of dplyr::mutate().

Let’s replicate a VLOOKUP in R using the new VLOOKUP() function. First, let’s create a simple lookup table.

lookup_table <- lookup_table <- tibble(
    stock   = c("FB", "AMZN", "NFLX", "GOOG"),
    company = c("Facebook", "Amazon", "Netflix", "Google")
)

lookup_table %>% kable()
stock company
FB Facebook
AMZN Amazon
NFLX Netflix
GOOG Google

Simple VLOOKUP Case

First, let’s mimic the “simple” case where we just want to lookup a Single Value.

VLOOKUP("AMZN", lookup_table, stock, company)
## [1] "Amazon"

So what happened? We supplied the string “AMZN”, and the VLOOKUP() function new to search the lookup_table matching the stock column and returning the company.

More Powerful VLOOKUP Case

Let’s try the more Powerful Case – pulling in a column of matched lookup values. We can do this by using the mutate() function from dplyr. This works because VLOOKUP() is vectorized.

FANG %>%
    mutate(company = VLOOKUP(symbol, lookup_table, stock, company))
## # A tibble: 4,032 x 9
##    symbol date        open  high   low close    volume adjusted company 
##                           
##  1 FB     2013-01-02  27.4  28.2  27.4  28    69846400     28   Facebook
##  2 FB     2013-01-03  27.9  28.5  27.6  27.8  63140600     27.8 Facebook
##  3 FB     2013-01-04  28.0  28.9  27.8  28.8  72715400     28.8 Facebook
##  4 FB     2013-01-07  28.7  29.8  28.6  29.4  83781800     29.4 Facebook
##  5 FB     2013-01-08  29.5  29.6  28.9  29.1  45871300     29.1 Facebook
##  6 FB     2013-01-09  29.7  30.6  29.5  30.6 104787700     30.6 Facebook
##  7 FB     2013-01-10  30.6  31.5  30.3  31.3  95316400     31.3 Facebook
##  8 FB     2013-01-11  31.3  32.0  31.1  31.7  89598000     31.7 Facebook
##  9 FB     2013-01-14  32.1  32.2  30.6  31.0  98892800     31.0 Facebook
## 10 FB     2013-01-15  30.6  31.7  29.9  30.1 173242600     30.1 Facebook
## # … with 4,022 more rows

What I’m Most Excited About Using VLOOKUP for

I actually can’t wait to use VLOOKUP() in Shiny apps. There are many times when I want the user to input a variable (a Key), and internally on the Shiny Server convert it to something more useful in a table (a Value). I’ll showcase this technique LIVE in Learning Lab 30 – Shiny + Tidyquant Apps (Register Here for FREE).

✅ Summarizing “IFS” Functions

One of the functions that many Excel Users (including myself) become accustomed to is filtering summations, which I refer to as “IFS” functions. These are very handy at quickly filtering on conditions while aggregating your data.

Excel Sum-If’s (and friends)

Excel has SUMIFS(), COUNTIFS(), AVERAGEIFS(), and several more Summarizing “IFS” Functions. Here’s how they work:

  1. Develop a Condition to summarize: Sum Sales when Region = “East”
  2. Use one or more Conditioning Column(s) to develop a logical cases (e.g. region = “East”)
  3. Use a Summarizing Column to aggregate (e.g. SUMIFS(sales))
  4. Put it together returning a single value: SUMIFS(sales, region, “=East”)

Excel SUMIFS

Excel SUMIFS is now in R as SUM_IFS()

R Sum-If’s (and friends)

R now has a full suite of “IFS” functions. We can test them to get the basics.

SUM_IFS()

Summarizing things.

SUM_IFS(x = 1:10, x > 5)
## [1] 40

COUNT_IFS()

Counting things.

COUNT_IFS(x = letters, str_detect(x, "[a-c]"))
## [1] 3

Used in tidyverse

Let’s use COUNT_IFS() to count how many times high trade volume occurs in 2015. We can accomplish this shockingly easily by combining summarise() and the COUNT_IFS() function.

FANG %>%
    group_by(symbol) %>%
    summarise(
        high_volume_in_2015 = COUNT_IFS(volume,
                                        year(date) == 2015,
                                        volume > quantile(volume, 0.75))
    )
## # A tibble: 4 x 2
##   symbol high_volume_in_2015
##                   
## 1 AMZN                    62
## 2 FB                      15
## 3 GOOG                    19
## 4 NFLX                    54

✅ 100+ New Excel Functions

100+ Excel-based statistical, date/date-time, and financial math functions have been ported to R. The things I’m most excited about are Business Calendar calculations:

  • Business Holiday Calendars – Solves a major pain point with business date calculations. Integrations with lubridate and timeDate have enabled Holiday Date Sequences to automate calculation of Net Work Days and working periods.

  • Financial Calculations – I’ve ported NPV, IRR, FV, PV, PMT, and RATE. Then I realized that there’s an amazing package called FinCal. The plan is to leverage FinCal going forward.

100 Excel Functions

100+ Excel Functions now in R

Business Calendars: Factoring in Business Holidays made Easy

Businesses rely on their ability to accurately predict revenue. A key driver is whether or not the business is open (shocker!). For a business that’s closed on weekends and standard business holidays, it’s now super easy to calculate something simple like NET_WORKDAYS().

Net Working Days Example

When run with just a start and end, it returns the number of days excluding weekends.

NET_WORKDAYS("2020-01-01", "2020-07-01") # 131 Skipping Weekends
## [1] 131

But what about holidays? We have a new function called HOLIDAY_SEQUENCE() to calculate the business holidays between two dates (thanks to timeDate!).

HOLIDAY_SEQUENCE("2020-01-01", "2020-07-01", calendar = "NYSE")
## [1] "2020-01-01" "2020-01-20" "2020-02-17" "2020-04-10" "2020-05-25"

Now we can simply remove these dates from the Net Workdays calculation. We get 126 days removing standard business holidays.

NET_WORKDAYS("2020-01-01", "2020-07-01",
             holidays = HOLIDAY_SEQUENCE("2020-01-01", "2020-07-01",
                                         calendar = "NYSE")) # 126 Skipping 5 NYSE Holidays
## [1] 126

✅ NEW Tidyverse Functionality

summarise_by_time() is a new time-based variant of dplyr::summarise() that allows collapsing time-series data by “second”, “minute”, “hour”, “day”, “week”, “month”, “quarter”, and “year”.

By Month

Here’s a quick example summarizing by "month".

FANG %>%
    group_by(symbol) %>%

    # Collapse from daily to FIRST value by month
    summarise_by_time(
        .date_var  = date,
        .by        = "month",
        adjusted   = FIRST(adjusted)
    )
## # A tibble: 192 x 3
## # Groups:   symbol [4]
##    symbol date       adjusted
##              
##  1 AMZN   2013-01-01     257.
##  2 AMZN   2013-02-01     265 
##  3 AMZN   2013-03-01     266.
##  4 AMZN   2013-04-01     262.
##  5 AMZN   2013-05-01     248.
##  6 AMZN   2013-06-01     267.
##  7 AMZN   2013-07-01     282.
##  8 AMZN   2013-08-01     306.
##  9 AMZN   2013-09-01     289.
## 10 AMZN   2013-10-01     321.
## # … with 182 more rows

By Year

The benefit is that I can quickly switch periods.

FANG %>%
    group_by(symbol) %>%
    summarise_by_time(
        .date_var  = date,
        .by        = "year",
        adjusted   = FIRST(adjusted)
    )
## # A tibble: 16 x 3
## # Groups:   symbol [4]
##    symbol date       adjusted
##              
##  1 AMZN   2013-01-01    257. 
##  2 AMZN   2014-01-01    398. 
##  3 AMZN   2015-01-01    309. 
##  4 AMZN   2016-01-01    637. 
##  5 FB     2013-01-01     28  
##  6 FB     2014-01-01     54.7
##  7 FB     2015-01-01     78.4
##  8 FB     2016-01-01    102. 
##  9 GOOG   2013-01-01    361. 
## 10 GOOG   2014-01-01    556. 
## 11 GOOG   2015-01-01    525. 
## 12 GOOG   2016-01-01    742. 
## 13 NFLX   2013-01-01     13.1
## 14 NFLX   2014-01-01     51.8
## 15 NFLX   2015-01-01     49.8
## 16 NFLX   2016-01-01    110.

What I’m Most Excited About

I can use summarise_by_time() in Shiny Apps to make common summarization plots. In fact, I’m teaching it: Learning Lab 30 – Shiny + Tidyquant for Finance Apps (Register Here, It’s Free)

Summarize by Time - Shiny App

Summarizing by Time in a Shiny App

📅 NEW API Integration (Implementation scheduled for March)

Coming Soon – Tingo API is a popular free and open source for stock prices, cryptocurrencies, and intraday feeds from the IEX (Investors Exchange). I’m planning integration via the riingo package.

Summary

There’s a ton to learn. So much that I couldn’t possibly go over all of the new features in tidyquant v1.0.0 in this article. And, most importantly, you haven’t seen tidyquant tackle some real messy business problems.

I have good news. In Learning Lab 30 – Shiny + Tidyquant for Finance Apps (Register Here, It’s Free), I’m going to be tackling some real financial data and showing how we can do really important things like:

  • Perform Portfolio Analysis
  • Use NEW Excel Features
  • Build Shiny Apps with Pivot Tables, VLOOKUPs and SUMIFS.

Experience Shiny + tidyquant
Financial Modeling App with Shiny & tidyquant

Learn how to make a Shiny Finance App using Shiny + tidyquant for financial modeling automation – FOR FREE. Plus, I’ll be showing off my tidyquant 1.0.0 NEW Excel in R Features and how they make it super easy to leverage Shiny. Registration is a no-brainer. Sign up here. 👇

Register for Learning Lab 30 Here

Registration closes March 11th (Day of the event).

Lab 30 – Shiny + Tidyquant starts in…

countdownmail.com


To leave a comment for the author, please follow the link and comment on their blog: business-science.io.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.

Search R-bloggers

Sponsors

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)