Introduction to DataExplorer

This document introduces the package DataExplorer, and shows how it can help you with different tasks throughout your data exploration process.

The remaining of this guide will be organized in accordance with the goals. As the package evolves, more content will be added.

Data

We will be using the nycflights13 datasets for this document. If you have not installed the package, please do the following:

install.packages("nycflights13")
library(nycflights13)

There are 5 datasets in this package:

airlines
airports
flights
planes
weather

If you want to quickly visualize the structure of all, you may do the following:

library(DataExplorer)
data_list <- list(airlines, airports, flights, planes, weather)
plot_str(data_list)

You may also try plot_str(data_list, type = "r") for a radial network.

Now let’s merge all tables together for a more robust dataset for later sections.

merge_airlines <- merge(flights, airlines, by = "carrier", all.x = TRUE)
merge_planes <- merge(merge_airlines, planes, by = "tailnum", all.x = TRUE, suffixes = c("_flights", "_planes"))
merge_airports_origin <- merge(merge_planes, airports, by.x = "origin", by.y = "faa", all.x = TRUE, suffixes = c("_carrier", "_origin"))
final_data <- merge(merge_airports_origin, airports, by.x = "dest", by.y = "faa", all.x = TRUE, suffixes = c("_origin", "_dest"))

Exploratory Data Analysis

Exploratory data analysis is the process to get to know your data, so that you can generate and test your hypothesis. Visualization techniques are usually applied.

To get introduced to your newly created dataset:

introduce(final_data)

rows	columns	discrete_columns	continuous_columns	all_missing_columns	total_missing_values	total_observations	memory_usage
336,776	42	16	26	0	809,170	14,144,592	94,557,176

Missing values

Real-world data is messy. After running the basic descriptive statistics, you might be interested in the missing data profile. You can simple use plot_missing function for this.

plot_missing(final_data)

You may also store the missing data profile with missing_data <- plot_missing(final_data) for additional analysis.

Distributions

Bar Charts

To visualize frequency distributions for all discrete features:

plot_bar(final_data)

## 5 columns ignored with more than 50 categories.
## dest: 105 categories
## tailnum: 4044 categories
## time_hour: 6936 categories
## model: 128 categories
## name: 102 categories

Upon closer inspection of manufacturer variable, it is not hard to identify the following duplications:

AIRBUS and AIRBUS INDUSTRIE
CANADAIR and CANADAIR LTD
MCDONNELL DOUGLAS, MCDONNELL DOUGLAS AIRCRAFT CO and MCDONNELL DOUGLAS CORPORATION

Let’s clean it up and look at the manufacturer distribution again:

final_data[which(final_data$manufacturer == "AIRBUS INDUSTRIE"),]$manufacturer <- "AIRBUS"
final_data[which(final_data$manufacturer == "CANADAIR LTD"),]$manufacturer <- "CANADAIR"
final_data[which(final_data$manufacturer %in% c("MCDONNELL DOUGLAS AIRCRAFT CO", "MCDONNELL DOUGLAS CORPORATION")),]$manufacturer <- "MCDONNELL DOUGLAS"

plot_bar(final_data$manufacturer)

Frequently, it is very beneficial to look at bivariate frequency distribution. For example, to look at discrete features by arr_delay:

plot_bar(final_data, with = "arr_delay")

## 5 columns ignored with more than 50 categories.
## dest: 105 categories
## tailnum: 4044 categories
## time_hour: 6936 categories
## model: 128 categories
## name: 102 categories

The resulting distribution looks quite different from the regular frequency distribution.

Histograms

To visualize distributions for all continuous features:

plot_histogram(final_data)

Immediately, you could observe that there are datetime features to be further treated, e.g., concatenating year, month and day to form date, and/or adding hour and minute to form datetime.

For the purpose of this vignette, I will not go deep into the analytical tasks. However, we should set flight to categorical, since that is the flight number with no mathematical meaning:

final_data$flight <- as.factor(final_data$flight)

Correlation Analysis

To visualize correlation heatmap for all features:

plot_correlation(final_data, maxcat = 5L, use = "pairwise.complete.obs")

## 11 features with more than 5 categories ignored!
## dest: 105 categories
## tailnum: 4044 categories
## carrier: 16 categories
## flight: 3844 categories
## time_hour: 6936 categories
## name_carrier: 16 categories
## manufacturer: 32 categories
## model: 128 categories
## engine: 7 categories
## name: 102 categories
## tzone_dest: 8 categories

You may also choose to visualize only discrete/continuous features with:

plot_correlation(final_data, type = "c", use = "pairwise.complete.obs")
plot_correlation(final_data, type = "d", use = "pairwise.complete.obs")

Principle Component Analysis

While you can always do plot_prcomp(na.omit(final_data)) directly, but PCA works better with cleaner data. To perform and visualize PCA on some selected features:

pca_df <- na.omit(final_data[, c("origin", "name_carrier", "type", "engine", "dep_delay", "arr_delay", "air_time", "month", "hour", "year_planes", "seats", "speed")])

plot_prcomp(pca_df)

Slicing & dicing

Often, slicing and dicing data in different ways could be crucial to your analysis, and yields insights quickly.

Boxplots

Suppose you would like to build a model to predict arrival delays, you may visualize the distribution of all continuous features based on arrival delays with a boxplot:

## Reduce data size for demo purpose
arr_delay_df <- final_data[, c("arr_delay", "month", "day", "hour", "minute", "dep_delay", "distance", "year_planes", "seats", "speed")]

## Call boxplot function
plot_boxplot(arr_delay_df, by = "arr_delay")

Among all the subtle changes in correlation with arrival delays, you could immediately spot that planes with 300+ seats tend to have much longer delays (16 ~ 21 hours). You may now drill down further to verify or generate more hypotheses.

Scatterplots

An alternative visualization is scatterplot. For example:

## Reduce data size for demo purpose
arr_delay_df2 <- final_data[sample.int(nrow(final_data), 1000), c("arr_delay", "dep_time", "dep_delay", "arr_time", "air_time", "distance", "year_planes", "seats", "speed")]

## Call scatterplot function
plot_scatterplot(arr_delay_df2, by = "arr_delay", size = 0.5)

Feature Engineering

Feature engineering is the process of creating new features from existing ones. Newly engineered features often generate valuable insights.

For functions in this section, it is preferred to use data.table objects as input, and they will be updated by reference. Otherwise, output object will be returned matching the input class.

Replace missing values

Missing values may have meanings for a feature. Other than imputation methods, we may also set them to some logical values. For example, for discrete features, we may want to group missing values to a new category. For continuous features, we may want to set missing values to a known number based on existing knowledge.

In DataExplorer, this can be done by set_missing. The function automatically matches the argument for either discrete or continuous features, i.e., if you specify a number, all missing continuous values will be set to that number. If you specify a string, all missing discrete values will be set to that string. If you supply both, both types will be set.

## Return data.frame
final_df <- set_missing(final_data, list(0L, "unknown"))
## Column [dep_time]: Set 8255 missing values to 0
## Column [dep_delay]: Set 8255 missing values to 0
## Column [arr_time]: Set 8713 missing values to 0
## Column [arr_delay]: Set 9430 missing values to 0
## Column [air_time]: Set 9430 missing values to 0
## Column [year_planes]: Set 57912 missing values to 0
## Column [engines]: Set 52606 missing values to 0
## Column [seats]: Set 52606 missing values to 0
## Column [speed]: Set 335813 missing values to 0
## Column [lat_dest]: Set 7602 missing values to 0
## Column [lon_dest]: Set 7602 missing values to 0
## Column [alt_dest]: Set 7602 missing values to 0
## Column [tz_dest]: Set 7602 missing values to 0
## Column [tailnum]: Set 2512 missing values to unknown
## Column [type]: Set 52606 missing values to unknown
## Column [manufacturer]: Set 52606 missing values to unknown
## Column [model]: Set 52606 missing values to unknown
## Column [engine]: Set 52606 missing values to unknown
## Column [name]: Set 7602 missing values to unknown
## Column [dst_dest]: Set 7602 missing values to unknown
## Column [tzone_dest]: Set 7602 missing values to unknown
plot_missing(final_df)


## Update data.table by reference
# library(data.table)
# final_dt <- data.table(final_data)
# set_missing(final_dt, list(0L, "unknown"))
# plot_missing(final_dt)

Group sparse categories

From the bar charts above, we observed a number of discrete features with sparse categorical distributions. Sometimes, we want to group low-frequency categories to a new bucket, or reduce the number of categories to a reasonable range. group_category will do the work.

Take manufacturer feature for example, suppose we want to group the long tail to another category. We could try with bottom 20% (by count) first:

group_category(data = final_data, feature = "manufacturer", threshold = 0.2)

##   manufacturer   cnt       pct   cum_pct
## 1       AIRBUS 88193 0.2618744 0.2618744
## 2       BOEING 82912 0.2461933 0.5080677
## 3      EMBRAER 66068 0.1961779 0.7042456

As we can see, manufacturer will be shrinked down to 4 categories, i.e., AIRBUS, BOEING, EMBRAER, and OTHER. If you like this threshold, you may specify update = TRUE to update the original dataset:

final_df <- group_category(data = final_data, feature = "manufacturer", threshold = 0.2, update = TRUE)
plot_bar(final_df$manufacturer)

Instead of shrinking categories by frequency, you may also group the categories by another continuous metric. For example, if you want to bucket the carrier with bottom 20% distance traveled, you may do the following:

group_category(data = final_data, feature = "name_carrier", threshold = 0.2, measure = "distance")

##             name_carrier      cnt       pct   cum_pct
## 1  United Air Lines Inc. 89705524 0.2561422 0.2561422
## 2   Delta Air Lines Inc. 59507317 0.1699153 0.4260575
## 3        JetBlue Airways 58384137 0.1667082 0.5927657
## 4 American Airlines Inc. 43864584 0.1252495 0.7180152

Similarly, if you like it, you may add update = TRUE to update the original dataset.

final_df <- group_category(data = final_data, feature = "name_carrier", threshold = 0.2, measure = "distance", update = TRUE)
plot_bar(final_df$name_carrier)

Dummify data (one hot encoding)

To transform the data into binary format (so that ML algorithms can pick it up), dummify will do the job. The function preserves original data structure, so that only eligible discrete features will be turned into binary format.

plot_str(
    list(
        "original" = final_data,
        "dummified" = dummify(final_data, maxcat = 5L)
    )
)

## 11 features with more than 5 categories ignored!
## dest: 105 categories
## tailnum: 4044 categories
## carrier: 16 categories
## flight: 3844 categories
## time_hour: 6936 categories
## name_carrier: 16 categories
## manufacturer: 32 categories
## model: 128 categories
## engine: 7 categories
## name: 102 categories
## tzone_dest: 8 categories

Note the maxcat argument. If a discrete feature has more categories than maxcat, it will not be dummified. As a result, it will be returned touched.

Drop features

After viewing the feature distribution, you often want to drop features that are insignificant. For example, features like dst_origin has only one value, and it doesn’t provide any valuable information. You can use drop_columns to quickly drop features. The function takes either names or column indices.

identical(
  drop_columns(final_data, c("dst_origin", "dst_dest", "tzone_dest")),
  drop_columns(final_data, c(34, 41, 42))
)

## [1] TRUE

Data Reporting

To organize all the data profiling statistics into a report, you may use the create_report() function. It will run most of the EDA functions and output a html file.

create_report(final_data)

To maximize the usage of this function, always supply a response variable (if applicable) to automate various bivariate analyses. For example,

create_report(final_data, y = "arr_delay")

You may also customize each individual section by passing their corresponding arguments as a list. They will later be passed to do.call to be invoked. The default config file is listed below. Simply copy and edit as necessary.

## Customize report configuration
config <- list(
  "introduce" = list(),
  "plot_str" = list(
    "type" = "diagonal",
    "fontSize" = 35,
    "width" = 1000,
    "margin" = list("left" = 350, "right" = 250)
  ),
  "plot_missing" = list(),
  "plot_histogram" = list(),
  "plot_bar" = list(),
  "plot_correlation" = list("use" = "pairwise.complete.obs"),
  "plot_prcomp" = list(),
  "plot_boxplot" = list(),
  "plot_scatterplot" = list()
)
## Create final report
create_report(final_data, y = "arr_delay", config = config)