Copyright (C) 2018-2020 Mohammad Akhlaghi mohammad@akhlaghi.org
Copyright (C) 2020 Raul Infante-Sainz infantesainz@gmail.com
See the end of the file for license conditions.
Maneage is a fully working template for doing reproducible research (or writing a reproducible paper) as defined in the link below. If the link below is not accessible at the time of reading, please see the appendix at the end of this file for a portion of its introduction. Some slides are also available to help demonstrate the concept implemented here.
Maneage is created with the aim of supporting reproducible research by
making it easy to start a project in this framework. As shown below, it is
very easy to customize Maneage for any particular (research) project and
expand it as it starts and evolves. It can be run with no modification (as
described in README.md) as a demonstration and customized for use in any
project as fully described below.
A project designed using Maneage will download and build all the necessary libraries and programs for working in a closed environment (highly independent of the host operating system) with fixed versions of the necessary dependencies. The tarballs for building the local environment are also collected in a separate repository. The final output of the project is a paper. Notice the last paragraph of the Acknowledgments where all the necessary software are mentioned with their versions.
Below, we start with a discussion of why Make was chosen as the high-level language/framework for project management and how to learn and master Make easily (and freely). The general architecture and design of the project is then discussed to help you navigate the files and their contents. This is followed by a checklist for the easy/fast customization of Maneage to your exciting research. We continue with some tips and guidelines on how to manage or extend your project as it grows based on our experiences with it so far. The main body concludes with a description of possible future improvements that are planned for Maneage (but not yet implemented). As discussed above, we end with a short introduction on the necessity of reproducible science in the appendix.
Please don't forget to share your thoughts, suggestions and criticisms. Maintaining and designing Maneage is itself a separate project, so please join us if you are interested. Once it is mature enough, we will describe it in a paper (written by all contributors) for a formal introduction to the community.
When batch processing is necessary (no manual intervention, as in a reproducible project), shell scripts are usually the first solution that come to mind. However, the inherent complexity and non-linearity of progress in a scientific project (where experimentation is key) make it hard to manage the script(s) as the project evolves. For example, a script will start from the top/start every time it is run. So if you have already completed 90% of a research project and want to run the remaining 10% that you have newly added, you have to run the whole script from the start again. Only then will you see the effects of the last new steps (to find possible errors, or better solutions and etc).
It is possible to manually ignore/comment parts of a script to only do a special part. However, such checks/comments will only add to the complexity of the script and will discourage you to play-with/change an already completed part of the project when an idea suddenly comes up. It is also prone to very serious bugs in the end (when trying to reproduce from scratch). Such bugs are very hard to notice during the work and frustrating to find in the end.
The Make paradigm, on the other hand, starts from the end: the final target. It builds a dependency tree internally, and finds where it should start each time the project is run. Therefore, in the scenario above, a researcher that has just added the final 10% of steps of her research to her Makefile, will only have to run those extra steps. With Make, it is also trivial to change the processing of any intermediate (already written) rule (or step) in the middle of an already written analysis: the next time Make is run, only rules that are affected by the changes/additions will be re-run, not the whole analysis/project.
This greatly speeds up the processing (enabling creative changes), while keeping all the dependencies clearly documented (as part of the Make language), and most importantly, enabling full reproducibility from scratch with no changes in the project code that was working during the research. This will allow robust results and let the scientists get to what they do best: experiment and be critical to the methods/analysis without having to waste energy and time on technical problems that come up as a result of that experimentation in scripts.
Since the dependencies are clearly demarcated in Make, it can identify independent steps and run them in parallel. This further speeds up the processing. Make was designed for this purpose. It is how huge projects like all Unix-like operating systems (including GNU/Linux or Mac OS operating systems) and their core components are built. Therefore, Make is a highly mature paradigm/system with robust and highly efficient implementations in various operating systems perfectly suited for a complex non-linear research project.
Make is a small language with the aim of defining rules containing targets, prerequisites and recipes. It comes with some nice features like functions or automatic-variables to greatly facilitate the management of text (filenames for example) or any of those constructs. For a more detailed (yet still general) introduction see the article on Wikipedia:
Make is a +40 year old software that is still evolving, therefore many implementations of Make exist. The only difference in them is some extra features over the standard definition (which is shared in all of them). Maneage is primarily written in GNU Make (which it installs itself, you don't have to have it on your system). GNU Make is the most common, most actively developed, and most advanced implementation. Just note that Maneage downloads, builds, internally installs, and uses its own dependencies (including GNU Make), so you don't have to have it installed before you try it out.
The GNU Make book/manual (links below) is arguably the best place to learn Make. It is an excellent and non-technical book to help get started (it is only non-technical in its first few chapters to get you started easily). It is freely available and always up to date with the current GNU Make release. It also clearly explains which features are specific to GNU Make and which are general in all implementations. So the first few chapters regarding the generalities are useful for all implementations.
The first link below points to the GNU Make manual in various formats and in the second, you can download it in PDF (which may be easier for a first time reading).
If you use GNU Make, you also have the whole GNU Make manual on the
command-line with the following command (you can come out of the "Info"
environment by pressing q).
info make
If you aren't familiar with the Info documentation format, we strongly
recommend running $ info info and reading along. In less than an hour,
you will become highly proficient in it (it is very simple and has a great
manual for itself). Info greatly simplifies your access (without taking
your hands off the keyboard!) to many manuals that are installed on your
system, allowing you to be much more efficient as you work. If you use the
GNU Emacs text editor (or any of its variants), you also have access to all
Info manuals while you are writing your projects (again, without taking
your hands off the keyboard!).
The list below shows some of the works that have already been published with (earlier versions of) Maneage. Previously it was simply called "Reproducible paper template". Note that Maneage is evolving, so some details may be different in them. The more recent ones can be used as a good working example.
Infante-Sainz et al. (2020, MNRAS, 491, 5317): The version controlled project source is available on GitLab and is also archived on Zenodo with all the necessary software tarballs: zenodo.3524937.
Akhlaghi (2019, IAU Symposium 355). The version controlled project source is available on GitLab and is also archived on Zenodo with all the necessary software tarballs: zenodo.3408481.
Section 7.3 of Bacon et al. (2017, A&A 608, A1): The version controlled project source is available on GitLab and a snapshot of the project along with all the necessary input datasets and outputs is available in zenodo.1164774.
Section 4 of Bacon et al. (2017, A&A, 608, A1): The version controlled project is available on GitLab and a snapshot of the project along with all the necessary input datasets is available in zenodo.1163746.
Akhlaghi & Ichikawa (2015, ApJS, 220, 1): The version controlled project is available on GitLab. This is the very first (and much less mature!) incarnation of Maneage: the history of Maneage started more than two years after this paper was published. It is a very rudimentary/initial implementation, thus it is only included here for historical reasons. However, the project source is complete, accurate and uploaded to arXiv along with the paper.
A paper to fully describe Maneage has been submitted. Until then, if you used it in your work, please cite the paper that implemented its first version: Akhlaghi & Ichikawa (2015, ApJS, 220, 1).
Also, when your paper is published, don't forget to add a notice in your own paper (in coordination with the publishing editor) that the paper is fully reproducible and possibly add a sentence or paragraph in the end of the paper shortly describing the concept. This will help spread the word and encourage other scientists to also manage and publish their projects in a reproducible manner.
In order to customize Maneage to your research, it is important to first
understand its architecture so you can navigate your way in the directories
and understand how to implement your research project within its framework:
where to add new files and which existing files to modify for what
purpose. But if this the first time you are using Maneage, before reading
this theoretical discussion, please run Maneage once from scratch without
any changes (described in README.md). You will see how it works (note that
the configure step builds all necessary software, so it can take long, but
you can continue reading while its working).
The project has two top-level directories: reproduce and
tex. reproduce hosts all the software building and analysis
steps. tex contains all the final paper's components to be compiled into
a PDF using LaTeX.
The reproduce directory has two sub-directories: software and
analysis. As the name says, the former contains all the instructions to
download, build and install (independent of the host operating system) the
necessary software (these are called by the ./project configure
command). The latter contains instructions on how to use those software to
do your project's analysis.
After it finishes, ./project configure will create the following symbolic
links in the project's top source directory: .build which points to the
top build directory and .local for easy access to the custom built
software installation directory. With these you can easily access the build
directory and project-specific software from your top source directory. For
example if you run .local/bin/ls you will be using the ls of Maneage,
which is probably different from your system's ls (run them both with
--version to check).
Once the project is configured for your system, ./project make will do
the basic preparations and run the project's analysis with the custom
version of software. The project script is just a wrapper, and with the
make argument, it will first call top-prepare.mk and top-make.mk
(both are in the reproduce/analysis/make directory).
In terms of organization, top-prepare.mk and top-make.mk have an
identical design, only minor differences. So, let's continue Maneage's
architecture with top-make.mk. Once you understand that, you'll clearly
understand top-prepare.mk also. These very high-level files are
relatively short and heavily commented so hopefully the descriptions in
each comment will be enough to understand the general details. As you read
this section, please also look at the contents of the mentioned files and
directories to fully understand what is going on.
Before starting to look into the top top-make.mk, it is important to
recall that Make defines dependencies by files. Therefore, the
input/prerequisite and output of every step/rule must be a file. Also
recall that Make will use the modification date of the prerequisite(s) and
target files to see if the target must be re-built or not. Therefore during
the processing, many intermediate files will be created (see the tips
section below on a good strategy to deal with large/huge files).
To keep the source and (intermediate) built files separate, the user must
define a top-level build directory variable (or $(BDIR)) to host all the
intermediate files (you defined it during ./project configure). This
directory doesn't need to be version controlled or even synchronized, or
backed-up in other servers: its contents are all products, and can be
easily re-created any time. As you define targets for your new rules, it is
thus important to place them all under sub-directories of $(BDIR). As
mentioned above, you always have fast access to this "build"-directory with
the .build symbolic link. Also, beware to never make any manual change
in the files of the build-directory, just delete them (so they are
re-built).
In this architecture, we have two types of Makefiles that are loaded into
the top Makefile: configuration-Makefiles (only independent
variables/configurations) and workhorse-Makefiles (Makefiles that
actually contain analysis/processing rules).
The configuration-Makefiles are those that satisfy these two wildcards:
reproduce/software/config/*.conf (for building the necessary software
when you run ./project configure) and reproduce/analysis/config/*.conf
(for the high-level analysis, when you run ./project make). These
Makefiles don't actually have any rules, they just have values for various
free parameters throughout the configuration or analysis. Open a few of
them to see for yourself. These Makefiles must only contain raw Make
variables (project configurations). By "raw" we mean that the Make
variables in these files must not depend on variables in any other
configuration-Makefile. This is because we don't want to assume any order
in reading them. It is also very important to not define any rule, or
other Make construct, in these configuration-Makefiles.
Following this rule-of-thumb enables you to set these configure-Makefiles as a prerequisite to any target that depends on their variable values. Therefore, if you change any of their values, all targets that depend on those values will be re-built. This is very convenient as your project scales up and gets more complex.
The workhorse-Makefiles are those satisfying this wildcard
reproduce/software/make/*.mk and reproduce/analysis/make/*.mk. They
contain the details of the processing steps (Makefiles containing
rules). Therefore, in this phase order is important, because the
prerequisites of most rules will be the targets of other rules that will be
defined prior to them (not a fixed name like paper.pdf). The lower-level
rules must be imported into Make before the higher-level ones.
All processing steps are assumed to ultimately (usually after many rules)
end up in some number, image, figure, or table that will be included in the
paper. The writing of these results into the final report/paper is managed
through separate LaTeX files that only contain macros (a name given to a
number/string to be used in the LaTeX source, which will be replaced when
compiling it to the final PDF). So the last target in a workhorse-Makefile
is a .tex file (with the same base-name as the Makefile, but in
$(BDIR)/tex/macros). As a result, if the targets in a workhorse-Makefile
aren't directly a prerequisite of other workhorse-Makefile targets, they
can be a prerequisite of that intermediate LaTeX macro file and thus be
called when necessary. Otherwise, they will be ignored by Make.
Maneage also has a mode to share the build directory between several
users of a Unix group (when working on large computer clusters). In this
scenario, each user can have their own cloned project source, but share the
large built files between each other. To do this, it is necessary for all
built files to give full permission to group members while not allowing any
other users access to the contents. Therefore the ./project configure and
./project make steps must be called with special conditions which are
managed in the --group option.
Let's see how this design is implemented. Please open and inspect
top-make.mk it as we go along here. The first step (un-commented line) is
to import the local configuration (your answers to the questions of
./project configure). They are defined in the configuration-Makefile
reproduce/software/config/LOCAL.conf which was also built by ./project
configure (based on the LOCAL.conf.in template of the same directory).
The next non-commented set of the top Makefile defines the ultimate
target of the whole project (paper.pdf). But to avoid mistakes, a sanity
check is necessary to see if Make is being run with the same group settings
as the configure script (for example when the project is configured for
group access using the ./for-group script, but Make isn't). Therefore we
use a Make conditional to define the all target based on the group
permissions.
Having defined the top/ultimate target, our next step is to include all the
other necessary Makefiles. However, order matters in the importing of
workhorse-Makefiles and each must also have a TeX macro file with the same
base name (without a suffix). Therefore, the next step in the top-level
Makefile is to define the makesrc variable to keep the base names
(without a .mk suffix) of the workhorse-Makefiles that must be imported,
in the proper order.
Finally, we import all the necessary remaining Makefiles: 1) All the
analysis configuration-Makefiles with a wildcard. 2) The software
configuration-Makefile that contains their version (just in case its
necessary). 3) All workhorse-Makefiles in the proper order using a Make
foreach loop.
In short, to keep things modular, readable and manageable, follow these recommendations: 1) Set clear-to-understand names for the configuration-Makefiles, and workhorse-Makefiles, 2) Only import other Makefiles from top Makefile. These will let you know/remember generally which step you are taking before or after another. Projects will scale up very fast. Thus if you don't start and continue with a clean and robust convention like this, in the end it will become very dirty and hard to manage/understand (even for yourself). As a general rule of thumb, break your rules into as many logically-similar but independent steps as possible.
The reproduce/analysis/make/paper.mk Makefile must be the final Makefile
that is included. This workhorse Makefile ends with the rule to build
paper.pdf (final target of the whole project). If you look in it, you
will notice that this Makefile starts with a rule to create
$(mtexdir)/project.tex (mtexdir is just a shorthand name for
$(BDIR)/tex/macros mentioned before). As you see, the only dependency of
$(mtexdir)/project.tex is $(mtexdir)/verify.tex (which is the last
analysis step: it verifies all the generated results). Therefore,
$(mtexdir)/project.tex is the connection between the
processing/analysis steps of the project, and the steps to build the final
PDF.
During the research, it often happens that you want to test a step that is not a prerequisite of any higher-level operation. In such cases, you can (temporarily) define that processing as a rule in the most relevant workhorse-Makefile and set its target as a prerequisite of its TeX macro. If your test gives a promising result and you want to include it in your research, set it as prerequisites to other rules and remove it from the list of prerequisites for TeX macro file. In fact, this is how a project is designed to grow in this framework.
While Git does an excellent job at keeping a history of the contents of files, it makes no effort in keeping the file meta data, and in particular the dates of files. Therefore when you checkout to a different branch, files that are re-written by Git will have a newer date than the other project files. However, file dates are important in the current design of Maneage: Make checks the dates of the prerequisite files and target files to see if the target should be re-built.
To fix this problem, for Maneage we use a forked version of
Metastore. Metastore use
a binary database file (which is called .file-metadata) to keep the
modification dates of all the files under version control. This file is
also under version control, but is hidden (because it shouldn't be modified
by hand). During the project's configuration, Maneage installs to Git hooks
to run Metastore 1) before making a commit to update its database with the
file dates in a branch, and 2) after doing a checkout, to reset the
file-dates after the checkout is complete and re-set the file dates back to
what they were.
In practice, Metastore should work almost fully invisibly within your
project. The only place you might notice its presence is that you'll see
.file-metadata in the list of modified/staged files (commonly after
merging your branches). Since its a binary file, Git also won't show you
the changed contents. In a merge, you can simply accept any changes with
git add -u. But if Git is telling you that it has changed without a merge
(for example if you started a commit, but canceled it in the middle), you
can just do git checkout .file-metadata and set it back to its original
state.
Based on the explanation above, some major design points you should have in mind are listed below.
Define new reproduce/analysis/make/XXXXXX.mk workhorse-Makefile(s)
with good and human-friendly name(s) replacing XXXXXX.
Add XXXXXX, as a new line, to the values in makesrc of the top-level
Makefile.
Do not use any constant numbers (or important names like filter names)
in the workhorse-Makefiles or paper's LaTeX source. Define such
constants as logically-grouped, separate configuration-Makefiles in
reproduce/analysis/config/XXXXX.conf. Then set this
configuration-Makefiles file as a prerequisite to any rule that uses
the variable defined in it.
Through any number of intermediate prerequisites, all processing steps
should end in (be a prerequisite of) $(mtexdir)/verify.tex (defined in
reproduce/analysis/make/verify.mk). $(mtexdir)/verify.tex is the sole
dependency of $(mtexdir)/project.tex, which is the bridge between the
processing steps and PDF-building steps of the project.
Take the following steps to fully customize Maneage for your research
project. After finishing the list, be sure to run ./project configure and
project make to see if everything works correctly. If you notice anything
missing or any in-correct part (probably a change that has not been
explained here), please let us know to correct it.
As described above, the concept of reproducibility (during a project) heavily relies on version control. Currently Maneage uses Git as its main version control system. If you are not already familiar with Git, please read the first three chapters of the ProGit book which provides a wonderful practical understanding of the basics. You can read later chapters as you get more advanced in later stages of your work.
Get this repository and its history (if you don't already have it):
Arguably the easiest way to start is to clone Maneage and prepare for
your customizations as shown below. After the cloning first you rename
the default origin remote server to specify that this is Maneage's
remote server. This will allow you to use the conventional origin
name for your own project as shown in the next steps. Second, you will
create and go into the conventional master branch to start
committing in your project later.
git clone https://git.maneage.org/project.git # Clone/copy the project and its history.
mv project my-project # Change the name to your project's name.
cd my-project # Go into the cloned directory.
git remote rename origin origin-maneage # Rename current/only remote to "origin-maneage".
git checkout -b master # Create and enter your own "master" branch.
pwd # Just to confirm where you are.
Prepare to build project: The ./project configure command of the
next step will build the different software packages within the
"build" directory (that you will specify). Nothing else on your system
will be touched. However, since it takes long, it is useful to see
what it is being built at every instant (its almost impossible to tell
from the torrent of commands that are produced!). So open another
terminal on your desktop and navigate to the same project directory
that you cloned (output of last command above). Then run the following
command. Once every second, this command will just print the date
(possibly followed by a non-existent directory notice). But as soon as
the next step starts building software, you'll see the names of
software get printed as they are being built. Once any software is
installed in the project build directory it will be removed. Again,
don't worry, nothing will be installed outside the build directory.
# On another terminal (go to top project source directory, last command above)
./project --check-config
Test Maneage: Before making any changes, it is important to test it
and see if everything works properly with the commands below. If there
is any problem in the ./project configure or ./project make steps,
please contact us to fix the problem before continuing. Since the
building of dependencies in configuration can take long, you can take
the next few steps (editing the files) while its working (they don't
affect the configuration). After ./project make is finished, open
paper.pdf. If it looks fine, you are ready to start customizing the
Maneage for your project. But before that, clean all the extra Maneage
outputs with make clean as shown below.
./project configure # Build the project's software environment (can take an hour or so).
./project make # Do the processing and build paper (just a simple demo).
# Open 'paper.pdf' and see if everything is ok.
Setup the remote: You can use any hosting
facility
that supports Git to keep an online copy of your project's version
controlled history. We recommend GitLab because
it is more ethical (although not
perfect),
and later you can also host GitLab on your own server. Anyway, create
an account in your favorite hosting facility (if you don't already
have one), and define a new project there. Please make sure the newly
created project is empty (some services ask to include a README in
a new project which is bad in this scenario, and will not allow you to
push to it). It will give you a URL (usually starting with git@ and
ending in .git), put this URL in place of XXXXXXXXXX in the first
command below. With the second command, "push" your master branch to
your origin remote, and (with the --set-upstream option) set them
to track/follow each other. However, the maneage branch is currently
tracking/following your origin-maneage remote (automatically set
when you cloned Maneage). So when pushing the maneage branch to your
origin remote, you shouldn't use --set-upstream. With the last
command, you can actually check this (which local and remote branches
are tracking each other).
git remote add origin XXXXXXXXXX # Newly created repo is now called 'origin'.
git push --set-upstream origin master # Push 'master' branch to 'origin' (with tracking).
git push origin maneage # Push 'maneage' branch to 'origin' (no tracking).
Title, short description and author: The title and basic
information of your project's output PDF paper should be added in
paper.tex. You should see the relevant place in the preamble (prior
to \begin{document}. After you are done, run the ./project make
command again to see your changes in the final PDF, and make sure that
your changes don't cause a crash in LaTeX. Of course, if you use a
different LaTeX package/style for managing the title and authors (in
particular a specific journal's style), please feel free to use it
your own methods after finishing this checklist and doing your first
commit.
Delete dummy parts: Maneage contains some parts that are only for the initial/test run, mainly as a demonstration of important steps, which you can use as a reference to use in your own project. But they not for any real analysis, so you should remove these parts as described below:
paper.tex: 1) Delete the text of the abstract (from
\includeabstract{ to \vspace{0.25cm}) and write your own (a
single sentence can be enough now, you can complete it later). 2)
Add some keywords under it in the keywords part. 3) Delete
everything between %% Start of main body. and %% End of main
body.. 4) Remove the notice in the "Acknowledgments" section (in
\new{}) and Acknowledge your funding sources (this can also be
done later). Just don't delete the existing acknowledgment
statement: Maneage is possible thanks to funding from several
grants. Since Maneage is being used in your work, it is necessary to
acknowledge them in your work also.
reproduce/analysis/make/top-make.mk: Delete the delete-me line
in the makesrc definition. Just make sure there is no empty line
between the download \ and verify \ lines (they should be
directly under each other).
reproduce/analysis/make/verify.mk: In the final recipe, under the
commented line Verify TeX macros, remove the full line that
contains delete-me, and set the value of s in the line for
download to XXXXX (any temporary string, you'll fix it in the
end of your project, when its complete).
Delete all delete-me* files in the following directories:
rm tex/src/delete-me*
rm reproduce/analysis/make/delete-me*
rm reproduce/analysis/config/delete-me*
Disable verification of outputs by removing the yes from
reproduce/analysis/config/verify-outputs.conf. Later, when you are
ready to submit your paper, or publish the dataset, activate
verification and make the proper corrections in this file (described
under the "Other basic customizations" section below). This is a
critical step and only takes a few minutes when your project is
finished. So DON'T FORGET to activate it in the end.
Re-make the project (after a cleaning) to see if you haven't introduced any errors.
./project make clean
./project make
Don't merge some files in future updates: As described below, you
can later update your infra-structure (for example to fix bugs) by
merging your master branch with maneage. For files that you have
created in your own branch, there will be no problem. However if you
modify an existing Maneage file for your project, next time its
updated on maneage you'll have an annoying conflict. The commands
below show how to fix this future problem. With them, you can
configure Git to ignore the changes in maneage for some of the files
you have already edited and deleted above (and will edit below). Note
that only the first echo command has a > (to write over the file),
the rest are >> (to append to it). If you want to avoid any other
set of files to be imported from Maneage into your project's branch,
you can follow a similar strategy. We recommend only doing it when you
encounter the same conflict in more than one merge and there is no
other change in that file. Also, don't add core Maneage Makefiles,
otherwise Maneage can break on the next run.
echo "paper.tex merge=ours" > .gitattributes
echo "tex/src/delete-me.mk merge=ours" >> .gitattributes
echo "tex/src/delete-me-demo.mk merge=ours" >> .gitattributes
echo "reproduce/analysis/make/delete-me.mk merge=ours" >> .gitattributes
echo "reproduce/software/config/TARGETS.conf merge=ours" >> .gitattributes
echo "reproduce/analysis/config/delete-me-num.conf merge=ours" >> .gitattributes
git add .gitattributes
Copyright and License notice: It is necessary that all the
"copyright-able" files in your project (those larger than 10 lines)
have a copyright and license notice. Please take a moment to look at
several existing files to see a few examples. The copyright notice is
usually close to the start of the file, it is the line starting with
Copyright (C) and containing a year and the author's name (like the
examples below). The License notice is a short description of the
copyright license, usually one or two paragraphs with a URL to the
full license. Don't forget to add these two notices to any new
file you add in your project (you can just copy-and-paste). When you
modify an existing Maneage file (which already has the notices), just
add a copyright notice in your name under the existing one(s), like
the line with capital letters below. To start with, add this line with
your name and email address to paper.tex,
tex/src/preamble-header.tex, reproduce/analysis/make/top-make.mk,
and generally, all the files you modified in the previous step.
Copyright (C) 2018-2020 Existing Name <existing@email.address>
Copyright (C) 2020 YOUR NAME <YOUR@EMAIL.ADDRESS>
Configure Git for fist time: If this is the first time you are
running Git on this system, then you have to configure it with some
basic information in order to have essential information in the commit
messages (ignore this step if you have already done it). Git will
include your name and e-mail address information in each commit. You
can also specify your favorite text editor for making the commit
(emacs, vim, nano, and etc.).
git config --global user.name "YourName YourSurname"
git config --global user.email your-email@example.com
git config --global core.editor nano
Your first commit: You have already made some small and basic
changes in the steps above and you are in your project's master
branch. So, you can officially make your first commit in your
project's history and push it. But before that, you need to make sure
that there are no problems in the project. This is a good habit to
always re-build the system before a commit to be sure it works as
expected.
git status # See which files you have changed.
git diff # Check the lines you have added/changed.
./project make # Make sure everything builds successfully.
git add -u # Put all tracked changes in staging area.
git status # Make sure everything is fine.
git diff --cached # Confirm all the changes that will be committed.
git commit # Your first commit: put a good description!
git push # Push your commit to your remote.
Start your exciting research: You are now ready to add flesh and blood to this raw skeleton by further modifying and adding your exciting research steps. You can use the "published works" section in the introduction (above) as some fully working models to learn from. Also, don't hesitate to contact us if you have any questions.
High-level software: Maneage installs all the software that your
project needs. You can specify which software your project needs in
reproduce/software/config/TARGETS.conf. The necessary software are
classified into two classes: 1) programs or libraries (usually written
in C/C++) which are run directly by the operating system. 2) Python
modules/libraries that are run within Python. By default
TARGETS.conf only has GNU Astronomy Utilities (Gnuastro) as one
scientific program and Astropy as one scientific Python module. Both
have many dependencies which will be installed into your project
during the configuration step. To see a list of software that are
currently ready to be built in Maneage, see
reproduce/software/config/versions.conf (which has their versions
also), the comments in TARGETS.conf describe how to use the software
name from versions.conf. Currently the raw pipeline just uses
Gnuastro to make the demonstration plots. Therefore if you don't need
Gnuastro, go through the analysis steps in reproduce/analysis and
remove all its use cases (clearly marked).
Input dataset: The input datasets are managed through the
reproduce/analysis/config/INPUTS.conf file. It is best to gather all
the information regarding all the input datasets into this one central
file. To ensure that the proper dataset is being downloaded and used
by the project, it is also recommended get an MD5
checksum of the file and include
that in INPUTS.conf so the project can check it automatically. The
preparation/downloading of the input datasets is done in
reproduce/analysis/make/download.mk. Have a look there to see how
these values are to be used. This information about the input datasets
is also used in the initial configure script (to inform the users),
so also modify that file. You can find all occurrences of the demo
dataset with the command below and replace it with your input's
dataset.
grep -ir wfpc2 ./*
README.md: Correct all the XXXXX place holders (name of your
project, your own name, address of your project's online/remote
repository, link to download dependencies and etc). Generally, read
over the text and update it where necessary to fit your project. Don't
forget that this is the first file that is displayed on your online
repository and also your colleagues will first be drawn to read this
file. Therefore, make it as easy as possible for them to start
with. Also check and update this file one last time when you are ready
to publish your project's paper/source.
Verify outputs: During the initial customization checklist, you
disabled verification. This is natural because during the project you
need to make changes all the time and its a waste of time to enable
verification every time. But at significant moments of the project
(for example before submission to a journal, or publication) it is
necessary. When you activate verification, before building the paper,
all the specified datasets will be compared with their respective
checksum and if any file's checksum is different from the one recorded
in the project, it will stop and print the problematic file and its
expected and calculated checksums. First set the value of
verify-outputs variable in
reproduce/analysis/config/verify-outputs.conf to yes. Then go to
reproduce/analysis/make/verify.mk. The verification of all the files
is only done in one recipe. First the files that go into the
plots/figures are checked, then the LaTeX macros. Validation of the
former (inputs to plots/figures) should be done manually. If its the
first time you are doing this, you can see two examples of the dummy
steps (with delete-me, you can use them if you like). These two
examples should be removed before you can run the project. For the
latter, you just have to update the checksums. The important thing to
consider is that a simple checksum can be problematic because some
file generators print their run-time date in the file (for example as
commented lines in a text table). When checking text files, this
Makefile already has this function:
verify-txt-no-comments-leading-space. As the name suggests, it will
remove comment lines and empty lines before calculating the MD5
checksum. For FITS formats (common in astronomy, fortunately there is
a DATASUM definition which will return the checksum independent of
the headers. You can use the provided function(s), or define one for
your special formats.
Feedback: As you use Maneage you will notice many things that if implemented from the start would have been very useful for your work. This can be in the actual scripting and architecture of Maneage, or useful implementation and usage tips, like those below. In any case, please share your thoughts and suggestions with us, so we can add them here for everyone's benefit.
Re-preparation: Automatic preparation is only run in the first run
of the project on a system, to re-do the preparation you have to use
the option below. Here is the reason for this: when its necessary, the
preparation process can be slow and will unnecessarily slow down the
whole project while the project is under development (focus is on the
analysis that is done after preparation). Because of this, preparation
will be done automatically for the first time that the project is run
(when .build/software/preparation-done.mk doesn't exist). After the
preparation process completes once, future runs of ./project make
will not do the preparation process anymore (will not call
top-prepare.mk). They will only call top-make.mk for the
analysis. To manually invoke the preparation process after the first
attempt, the ./project make script should be run with the
--prepare-redo option, or you can delete the special file above.
./project make --prepare-redo
Pre-publication: add notice on reproducibility**: Add a notice somewhere prominent in the first page within your paper, informing the reader that your research is fully reproducible. For example in the end of the abstract, or under the keywords with a title like "reproducible paper". This will encourage them to publish their own works in this manner also and also will help spread the word.
The following is a list of design points, tips, or recommendations that have been learned after some experience with this type of project management. Please don't hesitate to share any experience you gain after using it with us. In this way, we can add it here (with full giving credit) for the benefit of others.
Modularity: Modularity is the key to easy and clean growth of a project. So it is always best to break up a job into as many sub-components as reasonable. Here are some tips to stay modular.
Short recipes: if you see the recipe of a rule becoming more than a handful of lines which involve significant processing, it is probably a good sign that you should break up the rule into its main components. Try to only have one major processing step per rule.
Context-based (many) Makefiles: For maximum modularity, this design
allows easy inclusion of many Makefiles: in
reproduce/analysis/make/*.mk for analysis steps, and
reproduce/software/make/*.mk for building software. So keep the
rules for closely related parts of the processing in separate
Makefiles.
Descriptive names: Be very clear and descriptive with the naming of the files and the variables because a few months after the processing, it will be very hard to remember what each one was for. Also this helps others (your collaborators or other people reading the project source after it is published) to more easily understand your work and find their way around.
Naming convention: As the project grows, following a single standard
or convention in naming the files is very useful. Try best to use
multiple word filenames for anything that is non-trivial (separating
the words with a -). For example if you have a Makefile for
creating a catalog and another two for processing it under models A
and B, you can name them like this: catalog-create.mk,
catalog-model-a.mk and catalog-model-b.mk. In this way, when
listing the contents of reproduce/analysis/make to see all the
Makefiles, those related to the catalog will all be close to each
other and thus easily found. This also helps in auto-completions by
the shell or text editors like Emacs.
Source directories: If you need to add files in other languages for
example in shell, Python, AWK or C, keep the files in the same
language in a separate directory under reproduce/analysis, with the
appropriate name.
Configuration files: If your research uses special programs as part
of the processing, put all their configuration files in a devoted
directory (with the program's name) within
reproduce/software/config. Similar to the
reproduce/software/config/gnuastro directory (which is put in
Maneage as a demo in case you use GNU Astronomy Utilities). It is
much cleaner and readable (thus less buggy) to avoid mixing the
configuration files, even if there is no technical necessity.
Contents: It is good practice to follow the following recommendations on the contents of your files, whether they are source code for a program, Makefiles, scripts or configuration files (copyrights aren't necessary for the latter).
Copyright: Always start a file containing programming constructs
with a copyright statement like the ones that Maneage starts with
(for example in the top level Makefile).
Comments: Comments are vital for readability (by yourself in two months, or others). Describe everything you can about why you are doing something, how you are doing it, and what you expect the result to be. Write the comments as if it was what you would say to describe the variable, recipe or rule to a friend sitting beside you. When writing the project it is very tempting to just steam ahead with commands and codes, but be patient and write comments before the rules or recipes. This will also allow you to think more about what you should be doing. Also, in several months when you come back to the code, you will appreciate the effort of writing them. Just don't forget to also read and update the comment first if you later want to make changes to the code (variable, recipe or rule). As a general rule of thumb: first the comments, then the code.
File title: In general, it is good practice to start all files with a single line description of what that particular file does. If further information about the totality of the file is necessary, add it after a blank line. This will help a fast inspection where you don't care about the details, but just want to remember/see what that file is (generally) for. This information must of course be commented (its for a human), but this is kept separate from the general recommendation on comments, because this is a comment for the whole file, not each step within it.
Make programming: Here are some experiences that we have come to learn over the years in using Make and are useful/handy in research contexts.
Environment of each recipe: If you need to define a special
environment (or aliases, or scripts to run) for all the recipes in
your Makefiles, you can use a Bash startup file
reproduce/software/shell/bashrc.sh. This file is loaded before every
Make recipe is run, just like the .bashrc in your home directory is
loaded every time you start a new interactive, non-login terminal. See
the comments in that file for more.
Automatic variables: These are wonderful and very useful Make
constructs that greatly shrink the text, while helping in
read-ability, robustness (less bugs in typos for example) and
generalization. For example even when a rule only has one target or
one prerequisite, always use $@ instead of the target's name, $<
instead of the first prerequisite, $^ instead of the full list of
prerequisites and etc. You can see the full list of automatic
variables
here. If
you use GNU Make, you can also see this page on your command-line:
info make "automatic variables"
Debug: Since Make doesn't follow the common top-down paradigm, it
can be a little hard to get accustomed to why you get an error or
un-expected behavior. In such cases, run Make with the -d
option. With this option, Make prints a full list of exactly which
prerequisites are being checked for which targets. Looking
(patiently) through this output and searching for the faulty
file/step will clearly show you any mistake you might have made in
defining the targets or prerequisites.
Large files: If you are dealing with very large files (thus having
multiple copies of them for intermediate steps is not possible), one
solution is the following strategy (Also see the next item on "Fast
access to temporary files"). Set a small plain text file as the
actual target and delete the large file when it is no longer needed
by the project (in the last rule that needs it). Below is a simple
demonstration of doing this. In it, we use Gnuastro's Arithmetic
program to add all pixels of the input image with 2 and create
large1.fits. We then subtract 2 from large1.fits to create
large2.fits and delete large1.fits in the same rule (when its no
longer needed). We can later do the same with large2.fits when it
is no longer needed and so on.
large1.fits.txt: input.fits
astarithmetic $< 2 + --output=$(subst .txt,,$@)
echo "done" > $@
large2.fits.txt: large1.fits.txt
astarithmetic $(subst .txt,,$<) 2 - --output=$(subst .txt,,$@)
rm $(subst .txt,,$<)
echo "done" > $@
reproduce/analysis/make/initialize.mk. This
wrapper will replace $(subst .txt,,XXXXX). Therefore, it will be
possible to greatly simplify this repetitive statement and make the
code even more readable throughout the whole project.Fast access to temporary files: Most Unix-like operating systems
will give you a special shared-memory device (directory): on systems
using the GNU C Library (all GNU/Linux system), it is /dev/shm. The
contents of this directory are actually in your RAM, not in your
persistence storage like the HDD or SSD. Reading and writing from/to
the RAM is much faster than persistent storage, so if you have enough
RAM available, it can be very beneficial for large temporary files to
be put there. You can use the mktemp program to give the temporary
files a randomly-set name, and use text files as targets to keep that
name (as described in the item above under "Large files") for later
deletion. For example, see the minimal working example Makefile below
(which you can actually put in a Makefile and run if you have an
input.fits in the same directory, and Gnuastro is installed).
.ONESHELL:
.SHELLFLAGS = -ec
all: mean-std.txt
shm-maneage := /dev/shm/$(shell whoami)-maneage-XXXXXXXXXX
large1.txt: input.fits
out=$$(mktemp $(shm-maneage))
astarithmetic $< 2 + --output=$$out.fits
echo "$$out" > $@
large2.txt: large1.txt
input=$$(cat $<)
out=$$(mktemp $(shm-maneage))
astarithmetic $$input.fits 2 - --output=$$out.fits
rm $$input.fits $$input
echo "$$out" > $@
mean-std.txt: large2.txt
input=$$(cat $<)
aststatistics $$input.fits --mean --std > $@
rm $$input.fits $$input
shm-maneage) has no suffix. So you can add the suffix
corresponding to your desired format afterwards (for example
$$out.fits, or $$out.txt). But more importantly, when mktemp
sets the random name, it also checks if no file exists with that name
and creates a file with that exact name at that moment. So at the end
of each recipe above, you'll have two files in your /dev/shm, one
empty file with no suffix one with a suffix. The role of the file
without a suffix is just to ensure that the randomly set name will
not be used by other calls to mktemp (when running in parallel) and
it should be deleted with the file containing a suffix. This is the
reason behind the rm $$input.fits $$input command above: to make
sure that first the file with a suffix is deleted, then the core
random file (note that when working in parallel on powerful systems,
in the time between deleting two files of a single rm command, many
things can happen!). When using Maneage, you can put the definition
of shm-maneage in reproduce/analysis/make/initialize.mk to be
usable in all the different Makefiles of your analysis, and you won't
need the three lines above it. Finally, BE RESPONSIBLE: after you
are finished, be sure to clean up any possibly remaining files (due
to crashes in the processing while you are working), otherwise your
RAM may fill up very fast. You can do it easily with a command like
this on your command-line: rm -f /dev/shm/$(whoami)-*.Software tarballs and raw inputs: It is critically important to document the raw inputs to your project (software tarballs and raw input data):
Keep the source tarball of dependencies: After configuration
finishes, the .build/software/tarballs directory will contain all
the software tarballs that were necessary for your project. You can
mirror the contents of this directory to keep a backup of all the
software tarballs used in your project (possibly as another version
controlled repository) that is also published with your project. Note
that software web-pages are not written in stone and can suddenly go
offline or not be accessible in some conditions. This backup is thus
very important. If you intend to release your project in a place like
Zenodo, you can upload/keep all the necessary tarballs (and data)
there with your
project. zenodo.1163746 is
one example of how the data, Gnuastro (main software used) and all
major Gnuastro's dependencies have been uploaded with the project's
source. Just note that this is only possible for free and open-source
software.
Keep your input data: The input data is also critical to the project's reproducibility, so like the above for software, make sure you have a backup of them, or their persistent identifiers (PIDs).
Version control: Version control is a critical component of Maneage. Here are some tips to help in effectively using it.
Regular commits: It is important (and extremely useful) to have the history of your project under version control. So try to make commits regularly (after any meaningful change/step/result).
Keep Maneage up-to-date: In time, Maneage is going to become more and more mature and robust (thanks to your feedback and the feedback of other users). Bugs will be fixed and new/improved features will be added. So every once and a while, you can run the commands below to pull new work that is done in Maneage. If the changes are useful for your work, you can merge them with your project to benefit from them. Just pay very close attention to resolving possible conflicts which might happen in the merge (updated settings that you have customized in Maneage).
git checkout maneage
git pull # Get recent work in Maneage
git log XXXXXX..XXXXXX --reverse # Inspect new work (replace XXXXXXs with hashs mentioned in output of previous command).
git log --oneline --graph --decorate --all # General view of branches.
git checkout master # Go to your top working branch.
git merge maneage # Import all the work into master.
Adding Maneage to a fork of your project: As you and your colleagues
continue your project, it will be necessary to have separate
forks/clones of it. But when you clone your own project on a
different system, or a colleague clones it to collaborate with you,
the clone won't have the origin-maneage remote that you started the
project with. As shown in the previous item above, you need this
remote to be able to pull recent updates from Maneage. The steps
below will setup the origin-maneage remote, and a local maneage
branch to track it, on the new clone.
git remote add origin-maneage https://git.maneage.org/project.git
git fetch origin-maneage
git checkout -b maneage --track origin-maneage/maneage
Commit message: The commit message is a very important and useful
aspect of version control. To make the commit message useful for
others (or yourself, one year later), it is good to follow a
consistent style. Maneage already has a consistent formatting
(described below), which you can also follow in your project if you
like. You can see many examples by running git log in the maneage
branch. If you intend to push commits to Maneage, for the consistency
of Maneage, it is necessary to follow these guidelines. 1) No line
should be more than 75 characters (to enable easy reading of the
message when you run git log on the standard 80-character
terminal). 2) The first line is the title of the commit and should
summarize it (so git log --oneline can be useful). The title should
also not end with a point (., because its a short single sentence,
so a point is not necessary and only wastes space). 3) After the
title, leave an empty line and start the body of your message
(possibly containing many paragraphs). 4) Describe the context of
your commit (the problem it is trying to solve) as much as possible,
then go onto how you solved it. One suggestion is to start the main
body of your commit with "Until now ...", and continue describing the
problem in the first paragraph(s). Afterwards, start the next
paragraph with "With this commit ...".
Project outputs: During your research, it is possible to checkout a
specific commit and reproduce its results. However, the processing
can be time consuming. Therefore, it is useful to also keep track of
the final outputs of your project (at minimum, the paper's PDF) in
important points of history. However, keeping a snapshot of these
(most probably large volume) outputs in the main history of the
project can unreasonably bloat it. It is thus recommended to make a
separate Git repo to keep those files and keep your project's source
as small as possible. For example if your project is called
my-exciting-project, the name of the outputs repository can be
my-exciting-project-output. This enables easy sharing of the output
files with your co-authors (with necessary permissions) and not
having to bloat your email archive with extra attachments also (you
can just share the link to the online repo in your
communications). After the research is published, you can also
release the outputs repository, or you can just delete it if it is
too large or un-necessary (it was just for convenience, and fully
reproducible after all). For example Maneage's output is available
for demonstration in a
separate repository.
Full Git history in one file: When you are publishing your project
(for example to Zenodo for long term preservation), it is more
convenient to have the whole project's Git history into one file to
save with your datasets. After all, you can't be sure that your
current Git server (for example GitLab, Github, or Bitbucket) will be
active forever. While they are good for the immediate future, you
can't rely on them for archival purposes. Fortunately keeping your
whole history in one file is easy with Git using the following
commands. To learn more about it, run git help bundle.
my-project-git.bundle to a descriptive name of your
project):
git bundle create my-project-git.bundle --all
my-project-git.bundle anywhere. Later, if
you need to un-bundle it, you can use the following command.
git clone my-project-git.bundle
This is an evolving project and as time goes on, it will evolve and become more robust. Some of the most prominent issues we plan to implement in the future are listed below, please join us if you are interested.
It is important to have control of the environment of the project. Maneage
currently builds the higher-level programs (for example GNU Bash, GNU Make,
GNU AWK and domain-specific software) it needs, then sets PATH so the
analysis is done only with the project's built software. But currently the
configuration of each program is in the Makefile rules that build it. This
is not good because a change in the build configuration does not
automatically cause a re-build. Also, each separate project on a system
needs to have its own built tools (that can waste a lot of space).
A good solution is based on the Nix package manager: a separate file is present for each software, containing all the necessary info to build it (including its URL, its tarball MD5 hash, dependencies, configuration parameters, build steps and etc). Using this file, a script can automatically generate the Make rules to download, build and install program and its dependencies (along with the dependencies of those dependencies and etc).
All the software are installed in a "store". Each installed file (library or executable) is prefixed by a hash of this configuration (and the OS architecture) and the standard program name. For example (from the Nix webpage):
/nix/store/b6gvzjyb2pg0kjfwrjmg1vfhh54ad73z-firefox-33.1/
The important thing is that the "store" is not in the project's search path. After the complete installation of the software, symbolic links are made to populate each project's program and library search paths without a hash. This hash will be unique to that particular software and its particular configuration. So simply by searching for this hash in the installed directory, we can find the installed files of that software to generate the links.
This scenario has several advantages: 1) a change in a software's build configuration triggers a rebuild. 2) a single "store" can be used in many projects, thus saving space and configuration time for new projects (that commonly have large overlaps in lower-level programs).
In case the link above is not accessible at the time of reading, here is a copy of the introduction of that link, describing the necessity for a reproducible project like this (copied on February 7th, 2018):
The most important element of a "scientific" statement/result is the fact that others should be able to falsify it. The Tsunami of data that has engulfed astronomers in the last two decades, combined with faster processors and faster internet connections has made it much more easier to obtain a result. However, these factors have also increased the complexity of a scientific analysis, such that it is no longer possible to describe all the steps of an analysis in the published paper. Citing this difficulty, many authors suffice to describing the generalities of their analysis in their papers.
However, It is impossible to falsify (or even study) a result if you can't exactly reproduce it. The complexity of modern science makes it vitally important to exactly reproduce the final result. Because even a small deviation can be due to many different parts of an analysis. Nature is already a black box which we are trying so hard to comprehend. Not letting other scientists see the exact steps taken to reach a result, or not allowing them to modify it (do experiments on it) is a self-imposed black box, which only exacerbates our ignorance.
Other scientists should be able to reproduce, check and experiment on the results of anything that is to carry the "scientific" label. Any result that is not reproducible (due to incomplete information by the author) is not scientific: the readers have to have faith in the subjective experience of the authors in the very important choice of configuration values and order of operations: this is contrary to the scientific spirit.
This file is part of Maneage's core: https://git.maneage.org/project.git
Maneage is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
Maneage is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with Maneage. If not, see https://www.gnu.org/licenses/.