Looks good

study.org

 #+TITLE: Minimal working example of an experimental study written in Org mode
 #+AUTHOR: Marek Felšöci
 #+DATE: {{{time(%B %d\, %Y)}}}
+#+OPTIONS: toc:nil
 #+LaTeX_CLASS: article
-#+LaTeX_CLASS_OPTIONS: [a4paper, 11pt, twoside, table]
+#+LaTeX_CLASS_OPTIONS: [a4paper, 11pt]
 
 * Introduction
 :PROPERTIES:
@@ -31,7 +32,7 @@ different parts of the model.
 
 #+CAPTION: A =cylinder= mesh counting 20,000 vertices.
 #+NAME: short-pipe
-#+ATTR_LaTeX: :width .5\columnwidth :placement [H]
+#+ATTR_LaTeX: :width .5\columnwidth
 [[./figures/cylinder.png]]
 
 It is to note that the HMAT solver implements a data compression method which
@@ -119,7 +120,8 @@ constructing our software environment and realizing the experiments. To do this,
 run the command below in a terminal.
 
 #+BEGIN_SRC shell
-guix time-machine --commit=58786c82df700fb1f1c30022ee6ec0b2e7516e7b -- \
+guix time-machine --no-channel-files \
+                  --commit=58786c82df700fb1f1c30022ee6ec0b2e7516e7b -- \
      shell --pure git emacs emacs-org -- emacs --batch --no-init-file -l org \
      --eval '(progn (setq org-src-preserve-indentation t) (org-babel-tangle-file "study.org"))'
 #+END_SRC
@@ -155,7 +157,7 @@ guix time-machine -C channels.scm -- shell --pure -m manifest.scm -- \
 
 The results are post-processed using scripts written in the R language. The
 following commands shall produce the figures featured further on in the
-document. See Section [[#plot-times]] and [[#plot-ram]] for a more detailed
+document. See Section [[#plot-time]] and [[#plot-ram]] for a more detailed
 explanation.
 
 #+BEGIN_SRC shell
@@ -245,7 +247,7 @@ itself as well as many other packages and libraries.
  (channel
   (name 'guix)
   (url "https://git.savannah.gnu.org/git/guix.git")
-  (commit "58786c82df700fb1f1c30022ee6ec0b2e7516e7b"))
+  (commit "10f3dd0e9e06d71d1bc1615c6a60cc3aa1ad1ff4"))
 #+END_SRC
 
 In addition to the ='guix= channel, we will need an extra channel, the
@@ -256,7 +258,7 @@ In addition to the ='guix= channel, we will need an extra channel, the
   (name 'minichannel)
   (url
    "https://gitlab.inria.fr/tutorial-guix-hpc-workshop/software/minichannel")
-  (commit "22d134707af275389da71e3f26d1119a7b7ad090")))
+  (commit "aaa4002f4cc4fc3af31977ab17b655cbb14fe053")))
 #+END_SRC
 
 For each channel, we specify the commit of the associated repository to acquire.
@@ -282,7 +284,7 @@ Scheme language, just like the channels.
 In our environment, we need the =minisolver= package, for the experimental part.
 
 #+BEGIN_SRC scheme
-(specification->manifest
+(specifications->manifest
   (list 
    "minisolver"
 #+END_SRC
@@ -304,6 +306,19 @@ vector graphics.
    "emacs-org"
    "emacs-org-ref"
    "emacs-ess"
+   "texlive-scheme-basic"
+       "texlive-hyperref"
+       "texlive-geometry"
+       "texlive-ec"
+       "texlive-float"
+"texlive-latexmk"
+"texlive-wrapfig"
+"texlive-amsmath"
+"texlive-ulem"
+"texlive-capt-of"
+"texlive-biblatex"
+"texlive-biber"
+       "sed"
    "inkscape"
 #+END_SRC
 
@@ -312,12 +327,7 @@ especially for gathering the machine info in the bash script defined in Section.
 
 #+BEGIN_SRC scheme
    "bash"
-   "coreutils"
-   "sed"
-   "grep"
-   "which"
-   "lshw"
-   "util-linux"))
+   "coreutils"))
 #+END_SRC
 
 ** Benchmarks
@@ -327,11 +337,11 @@ especially for gathering the machine info in the bash script defined in Section.
 
 We benchmark =minisolver= on linear systems with $N$, the number of unknowns,
 in (5,000; 10,000; 15,000; 20,000) while consider either the /low/ or the /high/
-compression level. Table [[benchmarks-table]] lists all of the considered
+# compression level. Table [[benchmarks-table]] lists all of the considered
 benchmarks.
 
 #+NAME: get-table
-#+begin_src elisp :noweb yes
+#+begin_src elisp :noweb yes :exports none
 <<get-table-of-benchmarks>>
 #+end_src
 
@@ -476,11 +486,11 @@ with increasing size of the target linear system.
 #+CAPTION: systems of varying size and using different compression levels.
 #+NAME: results-time
 #+ATTR_LaTeX: :width 1\columnwidth
-[[./figures/results-time.svg]]
+[[./figures/results-time.pdf]]
 
 **** Plot                                                          :noexport:
 :PROPERTIES:
-:CUSTOM_ID: plot-times
+:CUSTOM_ID: plot-time
 :header-args: :tangle results-time.R
 :END:
 
@@ -532,7 +542,7 @@ geom_point(size = 2.5) +
 Follows the configuration of axis and color legend labels
 
 #+begin_src R
-scale_x_continuous(name = "# Unknowns (\U1D441)") +
+scale_x_continuous(name = "# Unknowns (N)") +
 scale_y_continuous(name = "Computation time [s]") +
 labs(color = "Compression level") +
 #+end_src
@@ -556,7 +566,7 @@ At the very end of the script, we produce the plot into a =*.svg= file while
 specifying the destination as well as the initial dimensions.
 
 #+begin_src R
-ggsave(file = "figures/results-time.svg", plot = plot, width = 5, height = 5)
+ggsave(file = "figures/results-time.pdf", plot = plot, width = 5, height = 5)
 #+end_src
 
 *** Peak RAM usage
@@ -575,16 +585,16 @@ with increasing size of the target linear system.
 #+CAPTION: varying size and using different compression levels.
 #+NAME: results-ram
 #+ATTR_LaTeX: :width 1\columnwidth
-[[./figures/results-ram.svg]]
+[[./figures/results-ram.pdf]]
 
-**** Plot
+**** Plot :noexport:
 :PROPERTIES:
 :CUSTOM_ID: plot-ram
 :header-args: :tangle results-ram.R
 :END:
 
 To plot Figure [[results-ram]], we use nearly the same R script as in the case
-of Figure [[results-time]] (see Section [[plot-time]]).
+of Figure [[results-time]] (see Section [[#plot-time]]).
 
 #+begin_src R 
 library(svglite)
@@ -602,7 +612,7 @@ plot <- ggplot(
 ) +
 geom_line() +
 geom_point(size = 2.5) +
-scale_x_continuous(name = "# Unknowns (\U1D441)") +
+scale_x_continuous(name = "# Unknowns (N)") +
 scale_y_continuous(name = "RAM usage peaks [MiB]") +
 labs(color = "Compression level") +
 scale_color_manual(
@@ -614,7 +624,7 @@ theme_bw()
 The destination file name changes too, of course.
 
 #+begin_src R
-ggsave(file = "figures/results-ram.svg", plot = plot, width = 5, height = 5)
+ggsave(file = "figures/results-ram.pdf", plot = plot, width = 5, height = 5)
 #+end_src
 
 * Conclusion
@@ -693,11 +703,11 @@ a custom appearance configuration.
 
 #+begin_src elisp
 (setq org-latex-packages-alist '())
-(add-to-list 'org-latex-packages-alist '("" "minted"))
-(setq org-latex-listings 'minted)
-(setq org-latex-minted-options
-        '(("linenos = false") ("mathescape") ("breaklines")
-          ("style = monokai")))
+;;(add-to-list 'org-latex-packages-alist '("" "minted"))
+;;(setq org-latex-listings 'minted)
+;;(setq org-latex-minted-options
+;;        '(("linenos = false") ("mathescape") ("breaklines")
+;;          ("style = monokai")))
 #+end_src
 
 The export to a LaTeX PDF triggers the inclusion of a certain number of LaTeX
@@ -705,8 +715,8 @@ packages by default. However, we need to include an extra package, i.e. the
 =svg= package, allowing us to use =*.svg= files in the manuscript.
 
 #+begin_src elisp
-(add-to-list 'org-latex-packages-alist
-             '("inkscapelatex = false, inkscapearea = page" "svg"))
+;;(add-to-list 'org-latex-packages-alist
+;;             '("inkscapelatex = false, inkscapearea = page" "svg"))
 #+end_src
 
 Eventually, we set up a list of publishing targets. Here, we have a single