I am trying to convert some figures from *.pdf to *.eps via Inkscape and Python. I resorted to the Inkscape Manual and Inkscape Wiki and came up with the code below. However when I run the script, no output is written into the working directory.
import os
import subprocess
# Change to working directory
os.chdir("C:/Users/Username/Figures")
# Iterate over all PDF figures
for figure in [i for i in os.listdir() if i[-4:]=='.pdf']:
subprocess.run([
"C:/Program Files/Inkscape/inkscape.exe",
figure, # Input figure
f"--export-filename={figure[:-4]}.eps", # Output figure
"--batch-process"
])
There is also a similar question on Stackoverflow. However, if I replace *.emf with *.eps in the answer the files cannot be read by LaTeX.
Update: Inkscape version 0.92.4 (5da689c313, 2019-01-14)
I am preparing a custom model to run on android phone using instructions from https://www.tensorflow.org/mobile/prepare_models
First i retrained the model on custom images using below command:
$ python tensorflow/examples/image_retraining/retrain.py --image_dir tensorflow/examples/image_retraining/my_images/ --learning_rate=0.0005 --testing_percentage=15 --validation_percentage=15 --train_batch_size=32 --validation_batch_size=-1 --flip_left_right True --random_scale=30 --random_brightness=30 --eval_step_interval=100 --how_many_training_steps=100 --tfhub_module https://tfhub.dev/google/imagenet/mobilenet_v2_100_224/feature_vector/1
and as next step, I tested the model using label_image.py which also works fine in predicting the input image. However, freeze_graph gives error
$ bazel-bin/tensorflow/python/tools/freeze_graph --input_graph=/tmp/output_graph.pb --output_graph=/tmp/frozen_graph.pb
However, I keep getting this error.
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xff in position
57: invalid start byte
I noticed that your --input_graph=/tmp/output_graph.pb. Is your graph written as binary file (as_text=False), instead of pbtxt? If so, you will need to pass the --input_binary=true flag to freeze_graph.
if you write your graph as a binary file using:
tf.train.write_graph(sess.graph_def, 'tarinGraph', 'train2.pbtxt', as_text=False)
then you will need to pass the --input_binary=true flag to freeze_graph.
I'm currently working on a mobilenet pre-trained network which I would like to re-train with a dataset which contains png images.
I call the retrain script as follow :
python scripts/retrain.py
--bottleneck_dir=tf_files/bottlenecks
--how_many_training_steps=200
--model_dir=tf_files/models/
--summaries_dir=tf_files/training_summaries/"mobilenet_0.50_224"
--output_graph=tf_files/retrained_graph.pb
--output_labels=tf_files/retrained_labels.txt
--architecture mobilenet_0.50_224
--image_dir=tf_files/data
It seems like the images needs to be jpg, is it any way to work with png images instead ?
Can confirm it doesn't work with png files. I have, however, written a bash script that when placed in the same directory as the subclasses of the dataset can convert the images to jpg.
first you need to install imagemagick package by:
sudo apt-get install imagemagick
then you can run this script:
#!/bin/bash
for d in */ ; do
cd "$d"
for p in * ; do
IFS='.' read -r -a array <<< "$p"
convert "$p" "${array[0]}".jpg
done
cd ..
done
edit:
retrain.py does have a list with valid extensions (line 151):
extensions = ['jpg', 'jpeg', 'JPG', 'JPEG']
I didn't try to add 'png' to the list though
I use gnuplot with epslatex option to generate figure files for plotting purposes (like here). Via this method you get 2 files corresponding to same figure, one tex file and one eps file. The figure information is in eps file and font information is in tex file. So my question is this :
Can I combine both font information and figure content to a single file like pdf / eps file ?
UPDATE : OK I forgot to mention one thing. Off course set terminal postscript eps will give me eps outputs, but it will not embed latex symbols in the plot as labels etc.
So I found a method which I got from Christoph's comment. Set terminal like set terminal epslatex 8 standalone and then finally after plotting do something like below:
set terminal epslatex color standalone
set output "file.tex"
set xrange [1:500]
set ylabel "Variance (\\AA\\textsuperscript{2})" # angstoms
set mxtics 4
plot "version1.dat" using 1:3 with linespoints pointinterval -5 pt 10 lt 1 lw 3 title 'label1' , \
"version1.dat" using 1:2 with linespoints pointinterval -5 pt 6 lt -1 lw 3 title 'label2';
unset output
# And now the important part (combine info to single file) :
set output # finish the current output file
system('latex file.tex && dvips file.dvi && ps2pdf file.ps')
system('mv file.ps file.eps')
unset terminal
reset
These steps do output tex file which is converted to dvi and ps file. And finally you rename the postscript file to eps. Now you have figure information and tex symbol information in single file. This eps file is accepted by latex files.
OK now why this works : Sorry I don't know the entire technical details. But this is working fine with me.
I'm trying to convert to grayscale this PDF: https://dl.dropboxusercontent.com/u/10351891/page-27.pdf
Ghostscript (v 9.10) with pdfwrite Device fails with a "Unable to convert color space to Gray, reverting strategy to LeaveColorUnchanged." message.
I'm able to convert it through an intermediary ps file (using gs, pdftops (v 0.24.3) or pdf2ps) but this convertion rasterize the whole PDF.
I tryed a lot of other things: normalize the PDF using qpdf (v 5.0.1) or pdftk (v 1.44), transform it to a svg file and back to a PDF via Inkscape (v 0.48.4)... nothing seems to work.
The only one solution I found (which is not suitable for me in production environment) is to use Preview on my Mac and apply a Quartz Gray Tone filter manually or with an Automator script.
Anyone find another working way to do it?
Or is it possible to normalize the PDF or fix the issue to prevent the Ghostscript message "Unable to convert color space..." or to force the color space in another way?
Thanks!
gs \
-sDEVICE=pdfwrite \
-sProcessColorModel=DeviceGray \
-sColorConversionStrategy=Gray \
-dOverrideICC \
-o out.pdf \
-f page-27.pdf
This command converts your file to grayscale (GS 9.10).
A bit late in the day, but the top answer doesn't work for me with a different file. The underlying problem appears to be old code in Ghostscript, for which there is a later version that is not enabled by default. More on that here: http://bugs.ghostscript.com/show_bug.cgi?id=694608
The page above also gives a command that works for me:
gs \
-sDEVICE=pdfwrite \
-dProcessColorModel=/DeviceGray \
-dColorConversionStrategy=/Gray \
-dPDFUseOldCMS=false \
-o out.pdf \
-f in.pdf
Use the most recent code (not yet released) and set ColorConversionStrategy=Gray
If you crack into the file, you'll find that most of the colors are determined through an RGB ICC based color space (look for 8 0 R to find all the references to this colorspace). Perhaps gs is complaining about that?
Who knows.
The take away is that converting a page from one colorspace to another without affecting the content is non-trivial in that you need to be able to render the page and trap all changes to the current color/colorspace and substitute an equivalent in the target space as well as convert all image XObjects in the wrong colorspace, which will require decoding the image data and re-encoding it in the target space, as well as all form XObjects, which will be a task similar to trying to convert the parent page since form XObjects (I think your doc has 4) also contain resources and a content stream of page marking operators (which may include more XObjects).
It's certainly doable, but the process is nearly the same as rendering but with some fairly special-purpose code.
very late response, but the following command should work :
convert -colorspace GRAY input.pdf input_gray.pdf
In Linux:
Install pdftk
apt-get install pdftk
Once you have installed pdftk, save the script as graypdf.sh with the following code
# convert pdf to grayscale, preserving metadata
# "AFAIK graphicx has no feature for manipulating colorspaces. " http://groups.google.com/group/latexusersgroup/browse_thread/thread/5ebbc3ff9978af05
# "> Is there an easy (or just standard) way with pdflatex to do a > conversion from color to grayscale when a PDF file is generated? No." ... "If you want to convert a multipage document then you better have pdftops from the xpdf suite installed because Ghostscript's pdf to ps doesn't produce nice Postscript." http://osdir.com/ml/tex.pdftex/2008-05/msg00006.html
# "Converting a color EPS to grayscale" - http://en.wikibooks.org/wiki/LaTeX/Importing_Graphics
# "\usepackage[monochrome]{color} .. I don't know of a neat automatic conversion to monochrome (there might be such a thing) although there was something in Tugboat a while back about mapping colors on the fly. I would probably make monochrome versions of the pictures, and name them consistently. Then conditionally load each one" http://newsgroups.derkeiler.com/Archive/Comp/comp.text.tex/2005-08/msg01864.html
# "Here comes optional.sty. By adding \usepackage{optional} ... \opt{color}{\includegraphics[width=0.4\textwidth]{intro/benzoCompounds_color}} \opt{grayscale}{\includegraphics[width=0.4\textwidth]{intro/benzoCompounds}} " - http://chem-bla-ics.blogspot.com/2008/01/my-phd-thesis-in-color-and-grayscale.html
# with gs:
# http://handyfloss.net/2008.09/making-a-pdf-grayscale-with-ghostscript/
# note - this strips metadata! so:
# http://etutorials.org/Linux+systems/pdf+hacks/Chapter+5.+Manipulating+PDF+Files/Hack+64+Get+and+Set+PDF+Metadata/
COLORFILENAME=$1
OVERWRITE=$2
FNAME=${COLORFILENAME%.pdf}
# NOTE: pdftk does not work with logical page numbers / pagination;
# gs kills it as well;
# so check for existence of 'pdfmarks' file in calling dir;
# if there, use it to correct gs logical pagination
# for example, see
# http://askubuntu.com/questions/32048/renumber-pages-of-a-pdf/65894#65894
PDFMARKS=
if [ -e pdfmarks ] ; then
PDFMARKS="pdfmarks"
echo "$PDFMARKS exists, using..."
# convert to gray pdf - this strips metadata!
gs -sOutputFile=$FNAME-gs-gray.pdf -sDEVICE=pdfwrite \
-sColorConversionStrategy=Gray -dProcessColorModel=/DeviceGray \
-dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH "$COLORFILENAME" "$PDFMARKS"
else # not really needed ?!
gs -sOutputFile=$FNAME-gs-gray.pdf -sDEVICE=pdfwrite \
-sColorConversionStrategy=Gray -dProcessColorModel=/DeviceGray \
-dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH "$COLORFILENAME"
fi
# dump metadata from original color pdf
## pdftk $COLORFILENAME dump_data output $FNAME.data.txt
# also: pdfinfo -meta $COLORFILENAME
# grep to avoid BookmarkTitle/Level/PageNumber:
pdftk $COLORFILENAME dump_data output | grep 'Info\|Pdf' > $FNAME.data.txt
# "pdftk can take a plain-text file of these same key/value pairs and update a PDF's Info dictionary to match. Currently, it does not update the PDF's XMP stream."
pdftk $FNAME-gs-gray.pdf update_info $FNAME.data.txt output $FNAME-gray.pdf
# (http://wiki.creativecommons.org/XMP_Implementations : Exempi ... allows reading/writing XMP metadata for various file formats, including PDF ... )
# clean up
rm $FNAME-gs-gray.pdf
rm $FNAME.data.txt
if [ "$OVERWRITE" == "y" ] ; then
echo "Overwriting $COLORFILENAME..."
mv $FNAME-gray.pdf $COLORFILENAME
fi
# BUT NOTE:
# Mixing TEX & PostScript : The GEX Model - http://www.tug.org/TUGboat/Articles/tb21-3/tb68kost.pdf
# VTEX is a (commercial) extended version of TEX, sold by MicroPress, Inc. Free versions of VTEX have recently been made available, that work under OS/2 and Linux. This paper describes GEX, a fast fully-integrated PostScript interpreter which functions as part of the VTEX code-generator. Unless specified otherwise, this article describes the functionality in the free- ware version of the VTEX compiler, as available on CTAN sites in systems/vtex.
# GEX is a graphics counterpart to TEX. .. Since GEX may exercise subtle influence on TEX (load fonts, or change TEX registers), GEX is op- tional in VTEX implementations: the default oper- ation of the program is with GEX off; it is enabled by a command-line switch.
# \includegraphics[width=1.3in, colorspace=grayscale 256]{macaw.jpg}
# http://mail.tug.org/texlive/Contents/live/texmf-dist/doc/generic/FAQ-en/html/FAQ-TeXsystems.html
# A free version of the commercial VTeX extended TeX system is available for use under Linux, which among other things specialises in direct production of PDF from (La)TeX input. Sadly, it���s no longer supported, and the ready-built images are made for use with a rather ancient Linux kernel.
# NOTE: another way to capture metadata; if converting via ghostscript:
# http://compgroups.net/comp.text.pdf/How-to-specify-metadata-using-Ghostscript
# first:
# grep -a 'Keywo' orig.pdf
# /Author(xxx)/Title(ttt)/Subject()/Creator(LaTeX)/Producer(pdfTeX-1.40.12)/Keywords(kkkk)
# then - copy this data in a file prologue.ini:
#/pdfmark where {pop} {userdict /pdfmark /cleartomark load put} ifelse
#[/Author(xxx)
#/Title(ttt)
#/Subject()
#/Creator(LaTeX with hyperref package + gs w/ prologue)
#/Producer(pdfTeX-1.40.12)
#/Keywords(kkkk)
#/DOCINFO pdfmark
#
# finally, call gs on the orig file,
# asking to process pdfmarks in prologue.ini:
# gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 \
# -dPDFSETTINGS=/screen -dNOPAUSE -dQUIET -dBATCH -dDOPDFMARKS \
# -sOutputFile=out.pdf in.pdf prologue.ini
# then the metadata will be in output too (which is stripped otherwise;
# note bookmarks are preserved, however).
give the file excecution permissions
chmod +x greypdf.sh
And execute it like this:
./greypdf.sh input.pdf
It will create a file input-gray.pdf in the same location than the initial file
gs -dQUIET -dBATCH -dNOPAUSE -r150 -sDEVICE=pdfwrite -sProcessColorModel=DeviceGray -sColorConversionStrategy=Gray -dOverrideICC -sOutputFile=output.pdf input.pdf
You can use something which I created. It gives you the option to choose the specific page numbers that you want to convert to grayscale. Handy if you don't want to grayscale the entire pdf. https://github.com/shoaibkhan94/PdfGrayscaler.