#이맥스 팁 Tips and Tricks: The Non-Geek Scientist

2021-11-17 Bibliography bib tips

(Pierre Hily-Blant 2023)

from Pierre Hily-Blant - Tips and Tricks: The Non-Geek Scientist

Emacs

Needless to say how powerful emacs is. However, it can quickly become criptic. The Practical Emacs Tutorial by Xah Lee is an excellent reference for those like me who already know something, but very little.

Make programming easy for a non-geek scientist

There are many tools to develop complex codes, so-called IDEs (Integrated Development Environment). However, I tend to try and use as little CPU and power resources as possible, and therefore, try to minimize the number of softwares. Actually, developping with emacs is more than sufficient: it is extremely efficient. I have been developping codes with more than 1e5 lines with emacs. However, I am not a professional code developper and therefore, the following may only be of interest for scientists.

복잡한 코드를 개발하기 위한 도구, 소위 IDE(통합 개발 환경)가 많이 있습니다. 하지만 저는 CPU 와 전력 리소스를 최대한 적게 사용하려고 노력하기 때문에 소프트웨어의 수를 최소화하려고 노력합니다. 사실 이맥스로 개발하는 것만으로도 충분하고 매우 효율적입니다. 저는 이맥으로 1e5 줄 이상의 코드를 개발해본 적이 있습니다. 그러나 저는 전문 코드 개발자가 아니므로 다음 내용은 과학자들에게만 흥미로울 수 있습니다.

Latex code folding via `outline` minor mode

Adding the following lines to my .emacs.d/init.el allows to circulate through the Latex file. This is based on this reference.

(add-hook 'prog-mode-hook 'outline-minor-mode)
(global-set-key [M-left] 'outline-hide-body)
(global-set-key [M-right] 'outline-show-all)
(global-set-key [M-up] 'outline-previous-heading)
(global-set-key [M-down] 'outline-next-heading)
(global-set-key [M-S-left] 'outline-hide-subtree)
(global-set-key [M-S-right] 'outline-show-subtree)

Fortran code

Aim: have something as simple as org-mode to circulate within a program (e.g. Fortran). After spending some time, and thanks to brilliant emacs users, I found a minimal, though, sufficient, setting that does not need any additional package. This applies to Fortran, Latex codes, etc.

In my .emacs.d/init.el, I added the following:

;; this is from: https://www.reddit.com/r/emacs/comments/e2u5n9/code_folding_with_outlineminormode/
(add-hook 'prog-mode-hook 'outline-minor-mode)
(defun my-outline-fortran ()
"Fold definitions in Fortran."
(setq outline-regexp
(rx (or
;; Module and interface blocks.
(group (group (* space)) (or "module" "interface"))
;; Procedures and type definitions.
(group
(group (* space))
(*? (or "pure " "impure " "elemental "))
(or "function" "subroutine" "interface" "type" "type,")
(group (+ space)))))))
(add-hook 'f90-mode-hook 'my-outline-fortran)
;; The following is needed if you did not define these keys somewhere else.
;; (global-set-key [M-left] outline-hide-body)
;; (global-set-key [M-right] 'outline-show-all)
;; (global-set-key [M-up] 'outline-previous-heading)
;; (global-set-key [M-down] 'outline-next-heading)
;; (global-set-key [M-S-left] 'outline-hide-subtree)
;; (global-set-key [M-S-right] 'outline-show-subtree)

Editing

show line number on the left in the window: M-x linum-mode
Disable fontification in math mode =M-x customize-variable font-latex-fontify-script =
Scrolling “a la EXCEL” with a top row that is always visible: the trick is to split into two buffers, with the upper one containing only one row: C-u 1 C-x 2

Faces

Check the face of the character below the cursor:
- M-x describe-face: the easiest
- C-u C-x = will give many information and a link towards the specific face. If you are in a console or are doing emacs -nw in a terminal, instead of clicking, put the cursor on the appropriate element and press RET.
Inspect the value of frame-background-mode and customize it if needed

Emacs/org-mode

Using the truely fantastic org-mode tool makes emacs the most powerful tool for my everyday life as a scientist and teacher. The org-mode saves you a huge amount of time when producing rich READMEs, web pages

Editing

Commenting: C-c ; to comment an entire entry at once
Narrowing: to isolate an entry, use C-x n s and C-x n w to get back to the full TREE.
Navigation
- C-c n/p: Next/Previous heading, regardless of level
- C-c f/b: Next/Previous heading, same level
- C-c u: to go backward to upper level
To display all links in plain text: M-x org-toggle-link-display
- if you want to keep this behavior permanently, simply add (setq org-descriptive-links nil) to your init file

Tables

Build a table from a tabular text: copy/paste the table, then select it with C-x C-x and then C-c |
Operating on columns
- swap two columns: ALT + left/right
- remove current column: ALT + SHIFT + left/right
- insert column to the right: ALT + SHIFT + right
Duplication
- SHIFT + ENTER duplicates the current field in the row below; overwrites the content of the field;
- Similarly for rows, with up and down
Clear cell content: C-c SPC
C-c - insert a ----- line below the cursor
#+TBLFM to perform operations on columns of a table; several ones can be used, one per line, to be placed right below the table; C-c C-c applies the formula; output format controlled as follows: #+TBLFM: $2=9*$1**-2.3;%4.3g.
To add a row counter: #+TBLFM: $1=@#. This would start from row 1. Can be adjusted.
Sorting: put the cursor in the column to be sorted, and type C-c ^ and chose a value from the menu in the bottom menu bar.
C-c } Toggle the display of row and column numbers for a table, using overlays.
C-c + will compute the sum of the column above.
@>$3=vsum(@2..@-1)
- @> means ’last row'
- $3 is the 3rd column
- vsum is a vector sum; may be used either w/ columns or rows
- @-1 means last row above the current location, hence penultimate;
See here or here (orgmode.org) for more, and even more.

Appearance

org-indent-mode: see here

Using org-agenda

General link here
C-c a to open org-agenda dialog box

Using code blocks

To prevent org-mode and babel to evaluate code blocks during export, put this at the top of the org document:
```
  #+PROPERTY: header-args :eval never-export
```

Exporting

To export: C-c C-e and then follow the menu
HTML export
- C-c C-e P f will export the file within a project
- INFOJS_OPT must be given once and for all and on a single line
- #+OPTIONS: num:t will disable the folding property of infojs
LATEX export: typical sequence to produce a PDF is C-c C-e l p

Oral presentations with `org-mode`

org-mode can make it easy to produce oral presentation documents. I recommend to explore two ways: one is the reveal.js export which produces HTML presentations; nice, but quickly criptic. The other is to use the foiltex package which will sound stone age to those who knew it from the 2000s; yet, this produces simple and focused, high quality, PDF documents.

The main, major limitation I could not really solve is to place several pictures per page. There is no simple way.

With `foiltex` package

An example of a PDF file produced with this method may be seen here.

The following lines are adapted from this reference. I added them to my ~/.emacs.d/myinit.org.

  ;; Foiltex with emacs org-mode
  ;; from https://github.com/tomfaulkenberry/orgFoils
  (require 'ox-latex)
  (add-to-list 'org-latex-classes
  '("foils"
  "\\documentclass{foils}"
  ("\\foilhead{\\color{MidnightBlue}\\bf\\huge\\raggedright\\rule{10pt}{3ex}\\hspace{1em} %s}" . "\\section{%s}")
  ("\\foilhead{\\color{RedOrange}\\Large\\raggedright\\rule{10pt}{2ex}\\hspace{1em} %s}" . "\\subsection{%s}")
  ("\\foilhead{\\color{Blue}\\large\\raggedright\\rule{1ex}{1ex}\\hspace{1em} %s}" . "\\subsubsection{%s}")
  )
  )

The following file, ~/.emacs.d/org-templates/header-foiltex.org contains useful global definitions:

  #+AUTHOR: Pierre Hily-Blant
  #+email: pierre.hily-blant@univ-grenoble-alpes.fr
  #+LaTeX_CLASS: foils
  # +LaTeX_CLASS_OPTIONS: [footrule, 17pt, Screen16to9]
  #+LaTeX_CLASS_OPTIONS: [footrule, 17pt, Screen4to3]
  #+LATEX_HEADER: \MyLogo{Redefine MyLogo in LATEX\_HEADER}
  #+LATEX_HEADER: \setlength{\parindent}{0cm}
  #+LATEX_HEADER: \usepackage{array}
  #+LATEX_HEADER: \usepackage[dvipsnames]{xcolor}
  #+LATEX_HEADER: \newcolumntype{C}[1]{>{\centering\arraybackslash}p{#1}}
  #+LATEX_HEADER: \renewcommand\labelitemi{$\rhd$}
  # +LATEX_HEADER: \renewcommand\labelitemi{$\square$}
  #+LATEX_HEADER: \renewcommand\labelitemii{$\bullet$}
  # +LATEX_HEADER: \renewcommand\labelitemii{{\normalfont\bfseries --}}
  #+LATEX_HEADER: \renewcommand\labelitemiii{{\normalfont\bfseries --}}
  #+LATEX_HEADER: \usepackage{natbib}
  #+LATEX_HEADER: \usepackage{pxfonts}
  #+LATEX_HEADER: \usepackage[version=3]{mhchem}
  #+LATEX_HEADER: \setlength{\foilheadskip}{-30pt}
  #+LATEX_HEADER: \setlength{\headheight}{-10pt}
  #+LATEX_HEADER: \setlength{\headsep}{-10pt}
  #+LATEX_HEADER: \setlength{\textheight}{\paperheight}
  #+LATEX_HEADER: \addtolength{\textheight}{-2in}
  #+LATEX_HEADER: \newcommand\stress[1]{{\color{red}#1}}
  #+LaTeX_HEADER: \hypersetup{linktoc = all, colorlinks = true, urlcolor = MidnightBlue, citecolor = MidnightBlue, linkcolor = orange}
  # +latex_header: \usepackage[colorlinks=true,citecolor=electricblue,linkcolor=MidnightBlue]{hyperref}
  #+OPTIONS: toc:nil
  #+BIND: org-latex-images-centered nil

In addition to this global definition file, I found it useful to define a local file, which I called header.org and located in the directory containing my org files. This header.org file gathers all options common to my various presentations, to ensure consistency throughout a entire Lecture.

  #+setupfile: ~/.emacs.d/org-templates/header-foiltex.org
  # +TITLE: Stars: formation, structure, evolution
  #+LATEX_HEADER: \MyLogo{Physical Fluid Dynamics}
  # +AUTHOR: Pierre Hily-Blant (University Grenoble Alpes/IPAG)
  # +email: pierre.hily-blant@univ-grenoble-alpes.fr
  #+LATEX_HEADER: \MyLogo{Stars: formation, structure, evolution}
  #+LATEX_HEADER: \input{general.tex}
  #+LATEX_HEADER: \input{../book/STELLAR_shortcuts.tex}

Finally, an example of org-mode file could be

  #+SETUPFILE: ./header-stellar.org
  #+title: Demo
  #+subtitle: From protostars to stars

  * Example with a single image.
    #+attr_latex: :center t :height .5\textheight
    [[./figs/TTauri-joy1945.png]]
  ** Example with two images side by side
     #+attr_latex: :align C{0.4\textwidth}C{0.55\textwidth}
     | [[./figs/ET-1msol-overview.png]] | [[./figs/ET-5Msun-iben1991-annot.png]] |
  *** A subsubsection
      - with some items
      - written the usual way


      \newpage
      *Start of a new page*
      - with some items or wathever, such as an equation
        \[
        f(x) = int_0^a h(u) du
        \]

Shell

I don’t know much about shell commands. Yet, with very few commands, many highly complex operations can be done very fast on large files. It is usually much more efficient to use shell commands from scripts (particularly python which is so slow) to work on files. In this regard, awk and sed are must-know commands.

Loops

Slicing
- You can slice the input using ${@:3} or ${@:3:8} and then loop over it. For instance, to print arguments starting from 3
```
    for i in ${@:3} ; do echo $i; done
```
  or to print 8 arguments starting from 3 (so, arguments 3 through 10)
```
    for i in ${@:3:8} ; do echo $i; done
```

While loop

  i=1; while test $i -le 60; do let n=8000+$i ; echo $n ; let i=$i+1; done

Managing Strings

Change a generic part of filenames
- e.g. change the ‘file1’ string into ’newfilename’ in all files called file1.*.
```
    rename file1 newfilename file1.*
```
  is equivalent to
```
    for f in file1.*; do mv "$f" "${f/file1/newfilename}"; done
```
  You can also use regular expressions:
```
    rename 's/titi/toto/' tutu*
    rename 'y/a-z/A-Z/' *
```

Substitutions

  a=toto.jpeg
  b=jpeg
  c=gif
  echo ${a/$b/$c}
  toto.gif
  $ echo ${a/jpeg/gif}
  toto.gif

Substitutions using grep

Simple substitutions

    grep -e "subroutine\|intent\|\ function\ " tasks/grid_class.f90
    grep -v ^.*\! lib/*f90 | grep [$]

To exclude lines beginning with 0 or more blanks, followed by “!”
```
    grep -v '^\s*\!' spinstate.f90 | wc -l
```

Same as above, also excluding empty lines:

    grep -v '^$\|^\s*\!' spinstate.f90 | wc -l

Extract a given line from a file
```
  sed "NUMq;d" input_file
```
where NUM is the line number. In a callable routine with variables:
```
  sed "${NUM}q;d" $file
```
Change the field delimiter
- The following example shows how to reformat a data table into a Latex table.
```
    awk '$1=$1' FS=" " OFS=" & " analyse.dat| sed 's/$/\\\\/'
```

Sed

Read and print a file until a PATTERN is matched:
```
  sed '/PATTERN/q' filename
```
will print up to and including the match
```
  sed '/PATTERN/Q' filename
```
print up to BUT NOT including the match. See also with Awk.

More about printing a range of a file:

  sed -n '1,// p'  # from the first line until pattern is matched
  sed -n '//,$ p'  # from pattern to end of file
  sed -n '//,// p'  # from pattern1 to pattern2

Typical example for Latex formatting:
```
  sed 's/$/\\/' toto.sh
```

Suppress carriage return

  sed ':a;N;$!ba;s/\n/ /g' < input > output

Awk

Useful resources here and here

If, then, else:

  awk '{if (NF==1) {print $1} else print $0}' < tdv.tex

Extract part of a file (see also with Sed)

  awk '{print} /pattern/ {exit}' filename #from beginning to line matching 'pattern', included
  awk '/pattern/ {exit} {print}' filename #from beginning to line matching 'pattern', excluded

For loops

  for ((xx=4;xx<13;xx+=1)) ; do awk -v x=$xx 'BEGIN {printf "%6.1f",x/10.}  ($5 == x/10.) {printf "& %12.2e &%10.2e",$6,$7} END {printf "\\\\"}' < grid_abinit.log ; echo ; done
  awk '($1 !~ /^!/) {print $0}'

  Print a substring:
   awk '{print $1,"=",substr($0,index($0,$2))}' < ~/.gnuplot.d/global.cfg

will print the first field, followed by =, and the remaining of the input line; for more info on substrings etc, look at this link.

Texify a data table:

  awk 'BEGIN {OFS=" & ";ORS="\\\\\n"} {$1=$1} 1' opacity_one.dat

Find

Find files older than a given file:
```
  find . ! -cnewer ./yourfile -print
```

Find files and tar them:

  find / -type f -mtime -7 | xargs tar -rf weekly_incremental.tar

Pipe and more

To send, line by line, the output of a command:

  grep "png" stellar-1.org | sed -e 's/\[//g' -e 's/\]//g' -e 's/\.\//\.\.\/figs\//' | xargs -n1 ls

See here for a detailed explanation of xargs

Managing files

Dealing with large tar files

pigz -dc target.tar.gz | tar tf -
pigz -dc target.tar.gz | tar xf - 'filename'

Change ^M from MAC to \n in UNIX:

with tr

  for file in `ls` ; do tr '\r' '\n' < $file > dummy.txt ; mv dummy.txt $file ; done

This can also be done using sed:
```
  sed -i 's/^M/\n/g' filename
```

Suppress carriage return

with sed

  sed ':a;N;$!ba;s/\n/ /g' < input > output

with tr
```
  tr -d '\n' < input
```

Columns

Append files by columns

  paste file1.dat file2.dat
  paste file?.dat

Reformat files with constant column widths

  for file in `ls *txt` ; do column -t $file > dummy.txt ; mv dummy.txt $file ; done

Extract columns 2-3 from fullgrid_20K.log

  cat fullgrid_20K.log | tr -s ' ' | cut -d' ' -f3-4 > test.dat

Extract columns 4-5 from fullgrid_20K.log and collate them right to test.dat

  cat fullgrid_20K.log | tr -s ' ' | cut -d' ' -f5-6 | paste test.dat - > test2.dat

Replace the first 10 columns of fdens_1e6.out by columns 3-END of fullgrid_20K.log

  cat fullgrid_20K.log | tr -s ' ' | cut -d' ' -f4- | paste - <(awk '($1 !~/^\#/) {print $0}' fdens_1e6.out | tr -s ' ' | cut -d ' ' -f 11-) | sed 's/\ /\ \ /g' > test2.dat

Change the field delimiter

One usual example is to transform an ASCII table of numbers into a LATEX table.
```
  awk '$1=$1' FS=" " OFS=" & " < input_file | sed 's/$/\\\\/'
```
FS and OFS are awk special variables which means Input Field separator and Output field separator respectively. $1=$1 actually does nothing. For awk, to change the delimiter, there should be some change in the data and hence this dummy assignment.

Miscellaneous

tar --atime-preserve -c -v -z -f presentations.tarz Presentations
a2ps -M A4 --column=1 --font-size=10 check.dat -o check.ps

System commands

This command will usually tell you the vendor and model of your card. Open xterminal or shell prompt and type the command:
```
  lspci
  lspci -v
  lspci -v | less
```
To check loaded keys (for eg ssh)
```
  ssh-add -L
  ssh-rsa AAAAB....
```

Resize an image with `convert`

$ identify -format "%wx%h\n" test.jpg
1280x623
$ convert test.jpg -resize 640x312\> titi.jpg
$ identify -format "%wx%h\n" titi.jpg
640x312

Gnuplot

Gnuplot is the greatest plotting I found, especially since v5. Possibilities are huge, scripts are compact, commands are sufficiently simple and consistent to be memorized. In addition, since v5, the export to high-quality plots, which was one of the reasons I left gnuplot for years, is now at a high level, especially through the use of the cairo libraries. A particularly useful resource is the book “Gnuplot in action”. Another one, on-line resource is http://www.gnuplotting.org

Plotting

Formatting

  set format y "%g"
  set format y "10^{%L}"

Grid: an example to plot a grid at major and minor x and y tics in a log scale.

  set mxtics 10
  set mytics 10
  set grid mxtics
  set grid mytics
  set grid lt 8 lw 2, lt 8 lw 1

set offsets: to add space left/right (top/bottom) to avoid having the first and last points on the border; however, However, the margin’s size is not the input value in units of the axis; instead the value is rounded to the next auto-generated tic: set autoscale xfix can be used to enforce that inner margins are in unis of the selected axis.
x-axis labels; xtic() allows to build the labels of the x-axis directly from the values in given column of the data file; the following command would plot column 11 × 1000 with the row number as x (column 0), and with x-axis labels corresponding to the concatenation of the columns 3 and 4. The labels are rotated by 60 degrees, centered, and shifted downwards by 1 character.

Scripting

The load command loads a gnuplot script file; no arguments
```
  load 'toto.gnu'
```

The call command allows up to 9 arguments; the syntax in the script is like this:

  filename = ARG1
  if (ARGC >= 2){tkin=ARG2}
  if (ARGC == 3) {set title ARG3 offset 0,-5 font ",20"}

and the calling sequence is

  call 'toto.gnu' "toto.dat" 10 "TOTO"

Using MACROS: this is a very useful, although somewhat limited, tool; the following files are located in $HOME/.gnuplot.d/:

put the following png.cfg file:

    fileout=ARG2
    if (ARGC == 0) {print "Usage: @PNG  []" ; exit(0)} else {fileout=ARG2}
    if (ARGC < 2) {fileout="test.png"} else {fileout=ARG2}
    sfont=sprintf("\"Carlito,%s\"",ARG1)
    print sfont
    set term push
    set term pngcairo \
        notransparent \
        enhanced \
        font @sfont \
        fontscale 1 \
        rounded \
        crop \
        size 800,800
    set output fileout
    repl
    set output
    print "Output: ",fileout
    set term pop

in the general file general.cfg:

    set macro
    PNG = "call 'png.cfg'"

now, from gnuplot command line, typing the folling command will produce toto.png file using Carlito,18 font type and size:
```
    @PNG 18 "toto.png"
```

Loops, iterations, etc

In the following example, we iteratively plot the weight of each species, between column 74 to 83, relative to their sum, and label each curve by the name of the species with the column number concatenated to it
```
  plot [1e3:*] [1e-10:*] for [j=74:83] 'output/fdens_speci.out' u 2:(column(j)/(sum[k=74:83] column(k))) t columnhead(j).j
```
Here, we iteratively plot the weight of species taken from a list relative to the sum of all species corresponding to columns 74 to 83, and put a legend accordingly.
```
  plot [1e3:*] [1e-10:*] for [s in "CO NH3 H2O"] 'output/fdens_speci.out' u 2:(column(s)/(sum[k=74:83] column(k))) t s
```

Just checking this works really:

  plot [1e3:*] [1e-10:*] for [s in "CO NH3 H2O"] 'output/fdens_speci.out' u 2:(column(s)/(sum[k=74:83] column(k))) t s, \
  '' u 2:($18/($75+$83+$74+$77+$96+$78)) every 10 w p t "Checking"

Here, we have overplotted the H2O species, with points, taken every 10 samples to make it easy to read. See the result below:

A loop over three files:
A script taken from the “Gnuplot in Action” book:

Operation on columns within a loop

  s14 = "N N2 CN NO NH3 HCN"
  s15 = "N. N.N CN. N.O N.H3 HCN."
  set border 3
  ratio(k)=column(word(s14,k))/column(word(s15,k))
  fac(x) = x==2?2:1 # used to multiply by 2 if N2
  plot for [k=1:words(s14)] 'grid_tkin_SS.out' u 3:(ratio(k)*fac(k))

Labelling

This example would put an opaque box underneath the legend, shifting leftwards the symbols towards the text by 3 character units.
```
  set key at graph 0.1,0.95,0 left Left width -3 height 1 box opaque
```

The following puts thick dots at given points and label them A to C (here, orl(x,y) is a user-defined function):

  roots = "267.9 286.0 306.8" # for S = 1365 W m-2
  names = "A B C"
  r2=252.27
  r3=317.95
  unset label
  do for [i=1:3] {
         x = word(roots,i)
         ts = 1.*x
         set label word(names,i) at x,orl(x,0.8) point ps 2.5 pt 7 right offset -1,1
         cc=sprintf("%4s%12.1f%8.1f%12.1f", word(names,i),ts,ts-273.15,orl(x,0.8))
         print cc
         }

Exporting plots

pdfcairo: produces a PDF file; but not cropped… And issues with the size of the symbols: the size in the PDF does not reflect that in the window.
export to PNG:

An alternative, which allows to control directly the DPI: export as SVG and use in-line command inkscape to export to lossless PNG:

  set term svg enhanced size 500,500
  set output 'test.svg' ; repl ; set output
  inkscape --export-area-drawing --export-png=test.png --export-dpi=300 --export-background-opacity=1 test.svg

See alse here for some useful details.

Exporting outputs

To write the values in ASCII format:

  set table "tsurf.tex" separator "&"
  plot [0.1:50] [10:1000] 'tsurf.dat' \
       u (sprintf("%4.2f",$1)) : (sprintf("%4.2f",$2)) : (sprintf("%6.2f",ts($1,0))) with table
  unset table

Installation

By default, Debian install does not include the +LIBREADLINE; use show version long in gnuplot to see the compilation options;
You have to install it manually…
The following is taken from here.

Download the source:

You need to have the deb-src in the /etc/apt/sources.list file;

then

    cd /usr/local/src/
    sudo apt-get source gnuplot
    sudo apt build-dep gnuplot
    cd /usr/local/src/gnuplot-*.*.*/debian/

Edit the file rules by adding:

    --with-readline=gnu --enable-history-file

Then go up and compile:
```
    cd ..
    sudo dpkg-buildpackage
```
I noticed some loops when compiling the doc (as it seems to be) with images which are missing. Just wait…

Then, install the package:

    cd ..
    sudo dpkg -i gnuplot-*.deb

Exclude gnuplot from the update:

    ➜  src sudo apt-mark hold gnuplot
    gnuplot set on hold.

Image/PDF files

The basic tools are from pdftk and imagemagick

Concatenate PDF files

  man pdftk
  pdftk A=in1.pdf B=in2.pdf cat A1-end B3-end output test.pdf

Clip an image:
```
  convert -trim input.png output.png
```

Crop and clip an image: the following would split an image with four panels.

  fig="input.png"
  ww=`identify -format "%[fx:w] %[fx:h]\n" $fig | awk '{print $1}'`
  hh=`identify -format "%[fx:w] %[fx:h]\n" $fig | awk '{print $2}'`
  mx=`expr $ww / 2`
  my=`expr $hh / 2`
  #echo $ww $mx $hh $my
  convert $fig -crop -$mx-$my -trim topleft.png
  convert $fig -crop +$mx-$my -trim topright.png
  convert $fig -crop -$mx+$my -trim bottomleft.png
  convert $fig -crop +$mx+$my -trim bottomright.png
  exit 0

Crop a PDF file:
```
  pdfcrop -clip input.pdf output.pdf
```

Git

References

wiki de l’ensimag
Working with branches (and this is what you want to do!), here are some references I found really useful:
- https://nvie.com/posts/a-successful-git-branching-model/
- https://ensiwiki.ensimag.fr/index.php?title=G%C3%A9rer_des_branches_avec_Git
- Overall scheme for project organization: nice overview
  - Accordingly, a project would be organized with several branches
    - master
    - dev
    - temporary_feature_branch (e.g. read_input_parameters)
    - temporary_release_branch
    - temporary_hotfix_branch

Basic commands

Advice: to be exectued every time you’re leaving a session, no matter whether this is finished or not. The repository will thus behave as a backup. Obviously, do this in the dev branch, or in the master branch if this is version 0.x.
```
  git br # shows in which branch you are currently
```
To see what has changed; you have usually two lists:
- The first lists the files registered in the repository that have been modified since the last commit.
- The second lists unregistered files

To make the commit:

  git status
  git add -u  # list of files you will commit
  # git add *.f90 # would do what you think
  git commit -m "a message" # to commit the changes
  git push

Note that, if this is the first commit, then you have to do:

  git push origin  # to initialise the new branch with your version of the code

To revert a local file to the version in the repository:
```
  git checkout HEAD
```

Create aliases

  $ git config --global alias.co checkout
  $ git config --global alias.br branch
  $ git config --global alias.st status

Working with branches

This is essential. And this is, for me, the incredible plus of git to make such complex things so easy.

Create a branch

  git co -b       # to create a new branch

To create the branch on the remote server:

  $ git push origin
  Total 0 (delta 0), reused 0 (delta 0)
  To git+ssh://git.renater.fr:2222/scmrepos/git/chemcollapse/chemcollapse.git
   * [new branch]      test -> test

Pushing files from a branch for the first time:

  ➜  chemcollapse git:(dev) ✗ git push
  fatal: The current branch dev has no upstream branch.
  To push the current branch and set the remote as upstream, use

  git push --set-upstream origin dev

Moving from one branch to another

  git br      # to see where you are
  git co branch_name # to move to another branch

Initiate a branch: the point to be understood is that you will create a branch starting from an existing ones. For example, creating a branch dev should be done like this:
```
  $ git br        # check that you are in master
  $ git co -b dev # create the branch dev
  $ git br        # chech that you are in dev
```
From this point, you are starting a dev branch from the current version of the master branch.

Example, with a branch dedicated to improve the reading of the input parameters of ALICO:

  $ git push --set-upstream origin read_input_parameters
  Counting objects: 22, done.
  Delta compression using up to 4 threads.
  Compressing objects: 100% (22/22), done.
  Writing objects: 100% (22/22), 22.47 KiB | 0 bytes/s, done.
  Total 22 (delta 12), reused 0 (delta 0)
  To ssh://git.renater.fr:2222/alico.git
  * [new branch]      read_input_parameters -> read_input_parameters
  Branch read_input_parameters set up to track remote branch read_input_parameters from origin.

The option set-upstream tells git to push from our local read_input_parameters branch the one with the same name in the repository. This needs to be done only for the first push. After this, a simple git push is sufficient.

To delete a local or remote branch

  git branch -d the_local_branch      # To delete a local branch:
  git push origin --delete the_remote_branch # To delete a remote branch:

Conflicts

Resolving conflicts: see here

Problem with branches

Your branch and 'origin/master' have diverged, # and have 1 and 1 different commit(s) each, respectively.
- Taken from this thread
- check if you need to update origin.
  - If origin is up-to-date, then some commits have been pushed to origin from another repo while you made your own commits locally:
    ... o ---- o ---- A ---- B origin/master (upstream work) \ C master (your work)
  - As can be seen: commit C is based on commit A because that was the latest work you had fetched from upstream at the time.
  - However, before trying to push back to origin, someone else pushed commit B. Development history has diverged into separate paths.
  - You can then merge or rebase. See Pro Git: Git Branching - Rebasing for details.
- Merging (not my prefered solution)
  - Use the git merge command:
```
      $ git merge origin/master
```
  - This tells Git to integrate the changes from origin/master into your work and create a merge commit. The graph of history now looks like this:
    ... o ---- o ---- A ---- B origin/master (upstream work) \ \ C ---- M master (your work)
  - The new merge, commit M, has two parents, each representing one path of development that led to the content stored in that commit. Note that the history behind M is now non-linear.
- Rebase (my prefered solution)
  - Use the git rebase command:
```
      $ git rebase origin/master
```
  - This tells Git to replay commit C (your work) as if you had based it on commit B instead of A. CVS and Subversion users routinely rebase their local changes on top of upstream work when they update before commit. Git just adds explicit separation between the commit and rebase steps.
  - The graph of history now looks like this: You can review the differences with a:
    ... o ---- o ---- A ---- B origin/master (upstream work) \ C' master (your work)
  - Commit C’ is a new commit created by the git rebase command. It is different from C in two ways:
    1. It has a different history: B instead of A;
    2. Its content accounts for changes in both B and C; it is the same as M from the merge example.
  - Note that the history behind C’ is still linear which makes it my prefered choice. This approach preserves the CVS-based workflow used previously and may ease the transition. An attempt to push C’ into our repository will work (assuming you have permissions and no one has pushed while you were rebasing).
  - The git pull command provides a shorthand way to fetch from origin and rebase local work on it:
```
      $ git pull --rebase
```
    This combines the above fetch and rebase steps into one command.

RENATER server

http://sourcesup.renater.fr/
https://services.renater.fr/sourcesup/git/comment_travailler_avec_git_en_local_sur_son_poste_de_travail
Gestion des clés
- remise à jour des clés SSH: voir ici et là

Global installation

Debian

apt
- Useful links
  - Obsolete webpage but still useful
  - A low-level but useful webpage
- update and upgrade
  - The following packages have been kept back
    - Means that some package could not be upgraded; this could be due to several reasons; usually, this is because of dependency issues; most common issues are
      - the new package relies on a package for which there is no update in the current distro
      - the new package rests upon new package not installed in the current distro
    - To fix the issue, try:
      sudo apt-get install the_problematic_package
How to know which distribution is installed?
```
  cat /etc/os-release
```
produces the following output:
PRETTY_NAME=“Debian GNU/Linux 10 (buster)” NAME=“Debian GNU/Linux” VERSION_ID=“10” VERSION=“10 (buster)” VERSION_CODENAME=buster ID=debian HOME_URL=" https://www.debian.org/" SUPPORT_URL=" https://www.debian.org/support" BUG_REPORT_URL=" https://bugs.debian.org/"

Gnome

DVD ripper: Handbrake

Calendar To set Monday as the first weekday ?

  sudo cp /usr/share/i18n/locales/en_US /usr/share/i18n/locales/en_US.bak
  sudo emacs -nw /usr/share/i18n/locales/en_US

To this file, add the following line:

  first_weekday 2

The logout and login.

Sound-Recorder:

Comment definir l’entree pour enregistrement
- open sound-recorder
- open PulseAudio Volume Control (PAVC)
- start recording in sound-recorder
- set the input from the menu un the PAVC

internal –> internal microphone monitor of internal –> carte son

Mounting

http://askubuntu.com/questions/175739/how-do-i-remount-a-filesystem-as-read-write

  The correct syntax is:

  sudo mount -o remount,rw /partition/identifier /mount/point

  The correct syntax is:

  sudo mount -o remount,rw /partition/identifier /mount/point

  Where mount/point is /partition/identifier's corresponding mountpoint,
  as listed by the following command:
  mount -v | grep "^/" | awk '{print "\nPartition identifier: " $1  "\n Mountpoint: "  $3}'

Bluetooth

ls /run/user/1000/gvfs/obex…

PDF viewer

qpdfview
- View toolbar: zoomIn,scaleFactor,zoomOut,separator,continuousMode,twoPagesMode,separator,rotateLeft,rotateRight,separator,fullscreen,presentation

Softwares, other

Python

Use python3 instead of python

Positional arguments: see this link
Example
- pos(author:”Oort, J”,2)papers which have “J. Oort” as the second author
- pos(author:”Oort, J”,1,3)papers which have “J. Oort” as first, second, or third author

WPS

Add a dictionnary

  sudo apt-get install hunspell-fr-comprehensive
  cd /opt/kingsoft/wps-office/office6/dicts/spellcheck
  sudo ln -s /usr/share/hunspell/fr_FR.aff main.aff
  sudo ln -s /usr/share/hunspell/fr_FR.dic main.dic
  cat /opt/kingsoft/wps-office/office6/dicts/spellcheck/fr_FR/dict.conf
  [Dictionary]
  DisplayName=French (Comprehensive)
  DisplayName[de_DE]=French (Comprehensive)

R1C1 reference style: Menu -> Options -> General and save

Libreoffice

To Start With a Defined Page Number:
1. Click into the 1st paragraph of the page you want to set the number
2. Format -> Paragraph -> Text flow.
3. In the Breaks area, enable Insert and With Page Style. Set the page number.

Gimp

Put a white background in a transparent png
- open png
- Layer -> New layer: chose backgroud layer
- The white layer is above the png file
- Layer > Stack -> Reverse layer order
- File -> Export…

Author: Pierre Hily-Blant ( pierre.hily-blant@univ-grenoble-alpes.fr)

Date:

Emacs 24.5.1 (Org mode 9.0.5)

Validate

Related-Notes

References

Pierre Hily-Blant. 2023. “Tips and Tricks: The Non-Geek Scientist.” 2023. https://ipag.osug.fr/~hilyblap/misc/tricks.html.

마지막 수정일자 2024-09-30

#데이터로그

#이맥스 팁 Tips and Tricks: The Non-Geek Scientist

Emacs

Make programming easy for a non-geek scientist

Latex code folding via outline minor mode

Fortran code

Editing

Faces

Emacs/org-mode

Editing

Tables

Appearance

Links

Using org-agenda

Using code blocks

Exporting

Oral presentations with org-mode

With foiltex package

Shell

Loops

Managing Strings

Sed

Awk

Find

Pipe and more

Managing files

Dealing with large tar files

Change ^M from MAC to \n in UNIX:

Suppress carriage return

Columns

Change the field delimiter

Miscellaneous

System commands

Resize an image with convert

Gnuplot

Plotting

Scripting

Loops, iterations, etc

Labelling

Exporting plots

Exporting outputs

Installation

Image/PDF files

Git

References

Basic commands

Working with branches

Conflicts

Problem with branches

RENATER server

Global installation

Debian

Gnome

Sound-Recorder:

Mounting

Bluetooth

PDF viewer

Softwares, other

Python

ADS

WPS

Libreoffice

Gimp

Related-Notes

References

Latex code folding via `outline` minor mode

Oral presentations with `org-mode`

With `foiltex` package

Resize an image with `convert`