R help - Jul 2015 - Trellis Plots: translating lattice xyplot() to ggplot() [RESOLVED]

On Fri, 10 Jul 2015, Yihui Xie wrote:
> Your LyX example has two problems:
Yihui, et al.:

   All fixed now. After adding the specific libraries the chuck calling
xyplot() to produce the graphic caused an error when I tried to generate a
dvips preview:

LaTeX Error: File 'figure/unnamed-chunk-7-1' not found.

I could not locate the file with any of these extensions:
.eps,.ps,.eps.gz,.ps.gz,.eps.Z

Which is what sent me down the path of specifying pdf() and dev.off().

   However, compiling with pdflatex produces the graphic in the document. All
part of being brand new to using knitr.

Thank you all very much,

Rich

I guess you didn't tell us you were compiling the document with dvips
(BTW, I'm surprised that dvips is still alive today...), otherwise the
solution would be simply to use the postscript device instead of the
default pdf device (i.e. use the chunk option dev='postscript').
Hopefully you learned some lessons on what a reproducible example
means.

Regards,
Yihui
--
Yihui Xie <xieyihui at gmail.com>
Web: http://yihui.name


On Sat, Jul 11, 2015 at 9:03 AM, Rich Shepard <rshepard at
appl-ecosys.com> wrote:
> On Fri, 10 Jul 2015, Yihui Xie wrote:
>
>> Your LyX example has two problems:
>
>
> Yihui, et al.:
>
>   All fixed now. After adding the specific libraries the chuck calling
> xyplot() to produce the graphic caused an error when I tried to generate a
> dvips preview:
>
> LaTeX Error: File 'figure/unnamed-chunk-7-1' not found.
>
> I could not locate the file with any of these extensions:
> .eps,.ps,.eps.gz,.ps.gz,.eps.Z
>
> Which is what sent me down the path of specifying pdf() and dev.off().
>
>   However, compiling with pdflatex produces the graphic in the document.
All
> part of being brand new to using knitr.
>
> Thank you all very much,
>
> Rich

On Sat, 11 Jul 2015, Yihui Xie wrote:
> I guess you didn't tell us you were compiling the document with dvips
   That's how I preview lyx docs for the past decade-and-a-half or so. The
final .pdf is generated with pdflatex.

   There is now an issue when a second figure is added to the document: it
won't compile.

   Attached are the .RData file from the cwd, the test.lyx file (lyx-2.1.3),
and the two test files from /tmp/.../lyx_tmpbuf3/.

   Here are the lyx messages when I tried compiling the document with
pdflatex:

16:51:11.255: Exporting ...
16:51:11.259: (buffer-export pdf2)
16:51:11.278: Rscript --verbose --no-save --no-restore
"/usr/share/lyx/scripts/lyxknitr.R"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.Rnw"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.tex"
ISO-8859-15
"/home/rshepard/documents/white-papers/water-chem-anal/"
16:51:11.281: running
16:51:11.286:   '/usr/lib/R/bin/R --slave --no-restore --no-save
--no-restore --file=/usr/share/lyx/scripts/lyxknitr.R --args
-- /tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/test.Rnw
-- / /tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpb
16:51:11.288: uf3/test.tex ISO-8859-15
/home/rshepard/documents/white-papers/water-chem-anal/'
16:51:11.292: 
16:51:11.697: Loading required package: methods
16:51:11.709: Loading required package: survival
16:51:11.712: Loading required package: graphics
16:51:11.721: Loading required package: stats
16:51:11.921: 
16:51:11.923: Attaching package: ?NADA?
16:51:11.925: 
16:51:11.927: The following object is masked from ?package:stats?:
16:51:11.929: 
16:51:11.932:     cor
16:51:11.933: 
16:51:11.969: 
16:51:11.970: Attaching package: ?zoo?
16:51:11.972: 
16:51:11.975: The following objects are masked from ?package:base?:
16:51:11.976: 
16:51:11.979:     as.Date, as.Date.numeric
16:51:11.981: 
16:51:12.091: 
16:51:12.092: 
16:51:12.094: processing file:
/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/test.Rnw
16:51:12.112:
   |
   |                                                                 |   0%
   | 
16:51:12.113:
   |....                                                             |   6%
16:51:12.116:   ordinary text without R code
16:51:12.118: 
16:51:12.120:
   |
   |........                                                         |  12%
16:51:12.123: label: unnamed-chunk-1 (with options) 
16:51:12.126: List of 1
16:51:12.128:  $ echo: logi FALSE
16:51:12.130: 
16:51:12.204:
   |
   |...........                                                      |  18%
16:51:12.205:   ordinary text without R code
16:51:12.207: 
16:51:12.210:
   |
   |...............                                                  |  24%
16:51:12.213: label: unnamed-chunk-2
16:51:12.239:
   |
   |...................                                              |  29%
16:51:12.240:   ordinary text without R code
16:51:12.242: 
16:51:12.244:
   |
   |.......................                                          |  35%
16:51:12.247: label: unnamed-chunk-3
16:51:12.293:
   |
   |...........................                                      |  41%
16:51:12.295:   ordinary text without R code
16:51:12.297: 
16:51:12.299:
   |
   |...............................                                  |  47%
16:51:12.302: label: unnamed-chunk-4
16:51:12.344:
   |
   |..................................                               |  53%
16:51:12.346:   ordinary text without R code
16:51:12.348: 
16:51:12.350:
   |
   |......................................                           |  59%
16:51:12.353: label: unnamed-chunk-5
16:51:12.432:
   |
   |..........................................                       |  65%
16:51:12.434:   ordinary text without R code
16:51:12.436: 
16:51:12.438:
   |
   |..............................................                   |  71%
16:51:12.441: label: unnamed-chunk-6
16:51:12.484:
   |
   |..................................................               |  76%
16:51:12.486:   ordinary text without R code
16:51:12.488: 
16:51:12.490:
   |
   |......................................................           |  82%
16:51:12.493: label: unnamed-chunk-7
16:51:12.837:
   |
   |.........................................................        |  88%
16:51:12.839:   ordinary text without R code
16:51:12.841: 
16:51:12.843:
   |
   |.............................................................    |  94%
16:51:12.846: label: unnamed-chunk-8
16:51:12.849: Quitting from lines 101-102
(/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/test.Rnw) 
16:51:12.851: Error in parse(text = x, srcfile = src) : 
16:51:12.853:   <text>:2:0: unexpected end of input
16:51:12.855: 1: xyplot(value ~ sampdate | variable, data=carlin.2.melt,
rm.na = T, ylab = 'Measured Value', xlab = 'Date'
16:51:12.857:    ^
16:51:12.859: Calls: knit ... <Anonymous> -> parse_all ->
parse_all.character -> parse
16:51:12.861: 
16:51:12.862: Execution halted
Systemcall.cpp (292): Systemcall: 'Rscript --verbose --no-save --no-restore
"/usr/share/lyx/scripts/lyxknitr.R"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.Rnw"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.tex"
ISO-8859-15
"/home/rshepard/documents/white-papers/water-chem-anal/"' finished
with exit
code 1
Error: Cannot convert file
----------------------------------------
An error occurred while running:
Rscript --verbose --no-save --no-restore
"/usr/share/lyx/scripts/lyxknitr.R"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.Rnw"
"/tmp/lyx_tmpdir.rGQZdOZ30938/lyx_tmpbuf3/""test.tex"
ISO-8859-15
"/home/rshepard/documents/white-papers/water-chem-anal/"
16:51:16.136: Error while exporting format: PDF (pdflatex)

   There is no log file in the tempbuf. Note the references to
unnamed-chunks. Each chunk is labeled with a number when opened in lyx; the
two figures each have a distinct label. I don't see the specific error
generated by adding a second figure to the document. If there are more files
needed to reproduce this situation, let me know and I'll provide it.

Rich
-------------- next part --------------
#LyX 2.1 created this file. For more info see http://www.lyx.org/
\lyxformat 474
\begin_document
\begin_header
\textclass scrreprt
\begin_preamble
\date{}
\usepackage{textcomp,url,multicol}
%\setkomafont{sectioning}{\rmfamily}
\end_preamble
\options abstract=on
\use_default_options false
\begin_modules
natbibapa
knitr
\end_modules
\maintain_unincluded_children false
\language english
\language_package default
\inputencoding auto
\fontencoding global
\font_roman palatino
\font_sans default
\font_typewriter default
\font_math auto
\font_default_family default
\use_non_tex_fonts false
\font_sc false
\font_osf false
\font_sf_scale 100
\font_tt_scale 100
\graphics default
\default_output_format default
\output_sync 0
\bibtex_command bibtex
\index_command default
\paperfontsize default
\spacing single
\use_hyperref false
\papersize letterpaper
\use_geometry false
\use_package amsmath 1
\use_package amssymb 1
\use_package cancel 1
\use_package esint 0
\use_package mathdots 1
\use_package mathtools 1
\use_package mhchem 1
\use_package stackrel 1
\use_package stmaryrd 1
\use_package undertilde 1
\cite_engine natbib
\cite_engine_type authoryear
\biblio_style humannat
\use_bibtopic false
\use_indices false
\paperorientation portrait
\suppress_date false
\justification true
\use_refstyle 0
\index Index
\shortcut idx
\color #008000
\end_index
\secnumdepth 2
\tocdepth 2
\paragraph_separation indent
\paragraph_indentation default
\quotes_language english
\papercolumns 1
\papersides 2
\paperpagestyle default
\tracking_changes false
\output_changes false
\html_math_output 0
\html_css_as_file 0
\html_be_strict false
\end_header

\begin_body

\begin_layout Standard
\begin_inset Flex Chunk
status open

\begin_layout Plain Layout

load('.RData')
\begin_inset Argument 1
status open

\begin_layout Plain Layout
echo=FALSE
\end_layout

\end_inset


\end_layout

\end_inset


\end_layout

\begin_layout Standard
Some text in here; separating knitr chunks.
\end_layout

\begin_layout Standard
\begin_inset Flex Chunk
status open

\begin_layout Plain Layout

carlin <- read.csv("./carlin.csv", header = TRUE, sep =
",", stringsAsFactors
 = F)
\end_layout

\end_inset


\end_layout

\begin_layout Standard
Next paragraph of text.
\end_layout

\begin_layout Standard
\begin_inset Flex Chunk
status collapsed

\begin_layout Plain Layout

carlin$sampdate <- as.Date(carlin$sampdate)
\end_layout

\begin_layout Plain Layout

str(carlin)
\end_layout

\end_inset


\end_layout

\begin_layout Standard
Another paragraph or two of text.
\end_layout

\begin_layout Standard
\begin_inset Flex Chunk
status open

\begin_layout Plain Layout

summary(carlin)
\end_layout

\end_inset

 
\end_layout

\begin_layout Standard
Plots of these distributions are the next step ...
\end_layout

\begin_layout Standard
\begin_inset Flex Chunk
status collapsed

\begin_layout Plain Layout

carlin.1 <- subset(carlin, select = siteid:CO3)
\end_layout

\begin_layout Plain Layout

carlin.2 <- subset(carlin, select = c(siteid, sampdate, Alk:Ca))
\end_layout

\begin_layout Plain Layout

carlin.3 <- subset(carlin, select = c(siteid, sampdate, Mg:Cr))
\end_layout

\begin_layout Plain Layout

carlin.4 <- subset(carlin, select = c(siteid, sampdate, Co:Hg))
\end_layout

\end_inset


\end_layout

\begin_layout Standard
To plot each chemical constituent's concentration as a function of
collection
 date the data format needs to be reshaped from wide to long:
\end_layout

\begin_layout Standard
\begin_inset Flex Chunk
status collapsed

\begin_layout Plain Layout

library(reshape2)
\end_layout

\begin_layout Plain Layout

carlin.1.melt <- melt(carlin.1, na.rm = F, id.vars = c('siteid',
'sampdate'))
\end_layout

\begin_layout Plain Layout

carlin.2.melt <- melt(carlin.2, na.rm = F, id.vars = c('siteid',
'sampdate'))
\end_layout

\begin_layout Plain Layout

carlin.3.melt <- melt(carlin.3, na.rm = F, id.vars = c('siteid',
'sampdate'))
\end_layout

\begin_layout Plain Layout

carlin.4.melt <- melt(carlin.4, na.rm = F, id.vars = c('siteid',
'sampdate'))
\end_layout

\end_inset


\end_layout

\begin_layout Standard
The scatter (X-Y) plots use only measured data; missing data are not included.
 These plots show the measured values over time for each measured constituent.
 For example, Figure 
\begin_inset CommandInset ref
LatexCommand ref
reference "fig: scatter-plot group 1"

\end_inset

 shows that most chemical constitent concentrations are low and have low
 variability.
\begin_inset Float figure
wide false
sideways false
status collapsed

\begin_layout Plain Layout
\align center
\begin_inset Flex Chunk
status collapsed

\begin_layout Plain Layout

library(lattice)
\end_layout

\begin_layout Plain Layout

xyplot(value ~ sampdate | variable, data=carlin.1.melt, rm.na = T, ylab =
'Measured
 Value', xlab = 'Date')
\end_layout

\end_inset


\end_layout

\begin_layout Plain Layout
\begin_inset Caption Standard

\begin_layout Plain Layout
\begin_inset CommandInset label
LatexCommand label
name "fig: scatter-plot group 1"

\end_inset

Variability of measured values in group 1 constituents over the period of
 record at the Carlin station.
\end_layout

\end_inset


\end_layout

\begin_layout Plain Layout

\end_layout

\end_inset

Because river discharge (in cubic feet per second) spiked at about 8,000
 cfs in the mis-1980s the Y-axis for all panels in the figure have the save
 range.
 This compresses low values yet any single constituent can be plotted by
 itself and the details of its variability over time easily seen.
\end_layout

\begin_layout Standard
The second group of constituents (Figure 
\begin_inset CommandInset ref
LatexCommand ref
reference "fig: scatter-plot group 2"

\end_inset

)
\begin_inset Float figure
wide false
sideways false
status collapsed

\begin_layout Plain Layout
\align center
\begin_inset Flex Chunk
status collapsed

\begin_layout Plain Layout

xyplot(value ~ sampdate | variable, data=carlin.2.melt, rm.na = T, ylab =
'Measured
 Value', xlab = 'Date'
\end_layout

\end_inset


\end_layout

\begin_layout Plain Layout
\begin_inset Caption Standard

\begin_layout Plain Layout
\begin_inset CommandInset label
LatexCommand label
name "fig: scatter-plot group 2"

\end_inset

Variability of measured values in group 2 constituents over the period of
 record at the Carlin station.
\end_layout

\end_inset


\end_layout

\end_inset

 reveals that nutrients (the nitrogen compounds, phosphate, and organic
 carbon) are all consistenly low with low variability while total dissolved
 solids (TDS) concentrations are very highly variable both within years
 (as shown by the different concentrations at the same date) and across
 the period of observations.
 Alkalinity and hardness are also quite variable at both time scales.
\end_layout

\end_body
\end_document
-------------- next part --------------
\batchmode
\makeatletter
\def\input at path{{/home/rshepard/documents/white-papers/water-chem-anal//}}
\makeatother
\documentclass[letterpaper,twoside,english,abstract=on]{scrreprt}
\usepackage{mathpazo}
\usepackage[T1]{fontenc}
\usepackage[latin9]{inputenc}

\makeatletter

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LyX specific LaTeX commands.
\pdfpageheight\paperheight
\pdfpagewidth\paperwidth


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Textclass specific LaTeX commands.
\usepackage[natbibapa]{apacite}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands.
\date{}
\usepackage{textcomp,url,multicol}
%\setkomafont{sectioning}{\rmfamily}

\makeatother

\usepackage{babel}
\begin{document}
<<echo=FALSE>>load('.RData')
@

Some text in here; separating knitr chunks.

<<>>carlin <- read.csv("./carlin.csv", header = TRUE,
sep = ",", stringsAsFactors = F)
@

Next paragraph of text.

<<>>carlin$sampdate <- as.Date(carlin$sampdate)
str(carlin)
@

Another paragraph or two of text.

<<>>summary(carlin)
@ 

Plots of these distributions are the next step ...

<<>>carlin.1 <- subset(carlin, select = siteid:CO3)
carlin.2 <- subset(carlin, select = c(siteid, sampdate, Alk:Ca))
carlin.3 <- subset(carlin, select = c(siteid, sampdate, Mg:Cr))
carlin.4 <- subset(carlin, select = c(siteid, sampdate, Co:Hg))
@

To plot each chemical constituent's concentration as a function of
collection date the data format needs to be reshaped from wide to
long:

<<>>library(reshape2)
carlin.1.melt <- melt(carlin.1, na.rm = F, id.vars = c('siteid',
'sampdate'))
carlin.2.melt <- melt(carlin.2, na.rm = F, id.vars = c('siteid',
'sampdate'))
carlin.3.melt <- melt(carlin.3, na.rm = F, id.vars = c('siteid',
'sampdate'))
carlin.4.melt <- melt(carlin.4, na.rm = F, id.vars = c('siteid',
'sampdate'))
@

The scatter (X-Y) plots use only measured data; missing data are not
included. These plots show the measured values over time for each
measured constituent. For example, Figure \ref{fig: scatter-plot group 1}
shows that most chemical constitent concentrations are low and have
low variability.
\begin{figure}
\begin{centering}
<<>>library(lattice)
xyplot(value ~ sampdate | variable, data=carlin.1.melt, rm.na = T, ylab =
'Measured Value', xlab = 'Date')
@
\par\end{centering}

\protect\caption{\label{fig: scatter-plot group 1}Variability of measured values
in
group 1 constituents over the period of record at the Carlin station.}


\end{figure}
Because river discharge (in cubic feet per second) spiked at about
8,000 cfs in the mis-1980s the Y-axis for all panels in the figure
have the save range. This compresses low values yet any single constituent
can be plotted by itself and the details of its variability over time
easily seen.

The second group of constituents (Figure \ref{fig: scatter-plot group 2})
\begin{figure}
\begin{centering}
<<>>xyplot(value ~ sampdate | variable, data=carlin.2.melt, rm.na =
T, ylab = 'Measured Value', xlab = 'Date'
@
\par\end{centering}

\protect\caption{\label{fig: scatter-plot group 2}Variability of measured values
in
group 2 constituents over the period of record at the Carlin station.}
\end{figure}
 reveals that nutrients (the nitrogen compounds, phosphate, and organic
carbon) are all consistenly low with low variability while total dissolved
solids (TDS) concentrations are very highly variable both within years
(as shown by the different concentrations at the same date) and across
the period of observations. Alkalinity and hardness are also quite
variable at both time scales.
\end{document}