Automated reporting: onepager Latex+R+knitr

automated_reporting_latex_r_knitr.Rnw

\documentclass{article}


\usepackage[margin=2cm,noheadfoot]{geometry}
%\usepackage{fancyhdr}
%\pagestyle{fancyplain}

\usepackage{eso-pic}


%\rhead{\includegraphics[height=5cm]{logo_dd}} % right logo
%\lhead{\includegraphics[height=3cm]{logo_dd}} % right logo
%\renewcommand{\headrulewidth}{0pt} % remove rule below header




\date{}
\pagenumbering{gobble}


\renewcommand\floatpagefraction{.9}
\renewcommand\dblfloatpagefraction{.9} % for two column documents
\renewcommand\topfraction{.9}
\renewcommand\dbltopfraction{.9} % for two column documents
\renewcommand\bottomfraction{.9}
\renewcommand\textfraction{.1}
\setcounter{totalnumber}{50}
\setcounter{topnumber}{50}
\setcounter{bottomnumber}{50}


<<init, echo = FALSE, include = FALSE>>=
library(nycflights13)
data(flights)
# data(planes)




library(dplyr)
library(ggplot2)
library(tidyr)
library(lubridate)
library(gridExtra)
library(xtable)
library(knitr)


# choose quarter
active_quarter <- 4

# add date, quarter (1-4), and (sorted) month abbreviations to dataframe
flights <- flights %>%
  mutate(date = ymd(paste(year, month, day, sep="-")),
         week = week(paste(date)),
         quarter = cut(month, breaks = 4, labels = c(1:4)),
        month_abb = factor(month.abb[flights$month], month.abb, ordered = TRUE)) %>%
  filter(quarter == active_quarter) %>%
  filter(week != 53) %>%
  na.omit() # delete rows with NAs

flights_active_quarter <- nrow(flights %>% filter(quarter == active_quarter))

flights_jfk <- nrow(flights %>% filter(origin == "JFK"))
flights_lga <- nrow(flights %>% filter(origin == "LGA"))
flights_ewr <- nrow(flights %>% filter(origin == "EWR"))

@




\newcommand\AtPagemyUpperLeft[1]{\AtPageLowerLeft{%
\put(\LenToUnit{0.8\paperwidth},\LenToUnit{0.8\paperheight}){#1}}}
\AddToShipoutPictureFG{
  \AtPagemyUpperLeft{{\includegraphics[width=5cm,keepaspectratio]{logo_dd}}}
}%



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\begin{document}



\title{\vspace{-2cm}Analysis of flights from NYC in 2013;\\
QUARTER: \Sexpr{active_quarter}}

\maketitle

\section*{Background and Objective}
This reports analyses the number of flights with origin one of the
airports of NYC.


\section*{Main figures}
In the present quarter (\textbf{Quarter: \Sexpr{active_quarter}}) there were a total of
\textbf{\Sexpr{flights_active_quarter}} flights. From \textbf{JFK} started
\textbf{\Sexpr{flights_jfk}} flights. From \textbf{LGA} started
\textbf{\Sexpr{flights_lga}} flights. From \textbf{EWR} started
\textbf{\Sexpr{flights_ewr}} flights.

Non eram nescius,
Brute, cum, quae summis ingeniis exquisitaque doctrina philosophi
Graeco sermone tractavissent, ea Latinis litteris mandaremus,
fore ut hic noster labor in varias reprehensiones incurreret.

<<figs_define, echo = FALSE>>=


# how many flights per origin airport and per month?
f_1 <- flights %>%
  filter(quarter == active_quarter) %>%
  select(origin, month_abb) %>%
  group_by(month_abb, origin) %>%
  summarise(flight_count = n())

p_f_5 <- f_1 %>%
  ggplot(aes(x = origin, y = flight_count)) + geom_bar(stat = "identity") +
  facet_wrap(~ month_abb) +
  # ggtitle("Number of flights per month from each NYC origin airport") +
  theme(axis.title.y = element_text(angle = 0)) +
  coord_flip() +
  scale_y_continuous(breaks = c(10000))


by_week <- flights %>%
  filter(quarter == active_quarter) %>%
  group_by(week=week(date), origin) %>%
  summarize(n=n(), delay=mean(arr_delay), n_dests=n_distinct(dest))


p_f_7 <- by_week %>%
  ggplot(aes(week, n, colour = origin)) + geom_line() +
  # ggtitle(paste("Number of flights per week of the year per origin airport; quarter", active_quarter)) +
  theme(axis.title.y = element_text(angle = 0))
@


\begin{figure}[h!]
\centering
<<figs_print, fig.height=2, out.width='1\\linewidth', echo = FALSE>>=
grid.arrange(p_f_5, p_f_7, ncol = 2)
@

\caption{Number of flights per month (left panel) and week (right panel)
from each NYC origin airport}
\end{figure}


<<compute_top10, echo = FALSE>>=
# top 10 delay flights
top_10_delayed_flights <-
flights %>%
  mutate(delay = arr_delay - dep_delay) %>%
  select(carrier, tailnum, flight, origin, dest, air_time, date, delay) %>%
  arrange(desc(delay)) %>%
  top_n(10, delay)

 top_delay <- max(top_10_delayed_flights$delay)

@

\section*{Top 10 delayed flights in last quarter}
In this quarter, the \textbf{maximum delay} was \textbf{\Sexpr{top_delay} minutes}.
nam quibusdam, et iis quidem non admodum indoctis, totum hoc displicet philosophari.
quidam autem non tam id reprehendunt, si remissius agatur, sed tantum studium
tamque multam operam ponendam in eo non arbitrantur.

<<print_table_top10,  echo = FALSE, results = "asis", warning = FALSE>>=
# kable(top_10_delayed_flights)
  options(xtable.comment = FALSE)
  suppressMessages(print(xtable(top_10_delayed_flights), type="latex"))
@



\end{document}

Use this code to create a automated data-based report. Latex is used for page layout; R for analysis. Knitr brings the folks together...