Invoking Sage#
To run Sage, you basically just need to type sage from the
command-line prompt to start the Sage interpreter. See the Sage
Installation Guide for information about making sure your
$PATH is set correctly, etc.
Command-line options for Sage#
SageMath version 10.3, Release Date: 2024-03-19
Running Sage, the most common options:
file.[sage|py|spyx] -- run given .sage, .py or .spyx file
-c cmd -- evaluate cmd as sage code. For example,
"sage -c 'print(factor(35))'" will
print "5 * 7".
Running Sage, other options:
--preparse file.sage -- preparse "file.sage", and produce
the corresponding Python file
"file.sage.py"
-q -- quiet; start with no banner
--min -- do not populate global namespace
(must be first option)
--nodotsage -- run Sage without using the user's
.sage directory: create and use a temporary
.sage directory instead.
--gthread, --qthread, --q4thread, --wthread, --pylab
-- pass the option through to IPython
--simple-prompt -- pass the option through to IPython: use
this option with sage-shell mode in emacs
--gdb -- run Sage under the control of gdb
--lldb -- run Sage under the control of lldb
Running external programs:
--cython [...] -- run Cython with the given arguments
--ecl [...], --lisp [...] -- run Sage's copy of ECL (Embeddable
Common Lisp) with the given arguments
--gap [...] -- run Sage's Gap with the given arguments
--gap3 [...] -- run Sage's Gap3 with the given arguments
--git [...] -- run Sage's Git with the given arguments
--gp [...] -- run Sage's PARI/GP calculator with the
given arguments
--ipython [...], --ipython3 [...]
-- run Sage's IPython using the default
environment (not Sage), passing additional
additional options to IPython
--jupyter [...] -- run Sage's Jupyter with given arguments
--kash [...] -- run Sage's Kash with the given arguments
(not installed currently, run sage -i kash)
--M2 [...] -- run Sage's Macaulay2 with the given arguments
--maxima [...] -- run Sage's Maxima with the given arguments
--mwrank [...] -- run Sage's mwrank with the given arguments
--pip [...] -- invoke pip, the Python package manager
--polymake [...] -- run Sage's Polymake with given arguments
--python [...], --python3 [...]
-- run the Python 3 interpreter
-R [...] -- run Sage's R with the given arguments
--singular [...] -- run Sage's singular with the given arguments
--sqlite3 [...] -- run Sage's sqlite3 with given arguments
Running the notebook:
-n [...], --notebook=[...]
-- start the notebook; valid options include
'default', 'jupyter', 'jupyterlab', and 'export'.
Current default is 'jupyter'.
Run "sage --notebook --help" for more details.
Testing files:
-t [options] <files|dir> -- test examples in .py, .pyx, .sage
or .tex files. Options:
--long -- include lines with the phrase 'long time'
--verbose -- print debugging output during the test
--all -- test all files
--optional -- also test all examples labeled "# optional"
--only-optional[=tags]
-- if no 'tags' are specified, only run
blocks of tests containing a line labeled
"# optional". If a comma-separated
list of tags is specified, only run block
containing a line labeled "# optional tag"
for any of the tags given, and in these blocks
only run the lines which are unlabeled or
labeled "# optional" or labeled
"# optional tag" for any of the tags given.
--randorder[=seed] -- randomize order of tests
--random-seed[=seed] -- random seed (integer) for fuzzing doctests
--new -- only test files modified since last commit
--initial -- only show the first failure per block
--debug -- drop into PDB after an unexpected error
--failed -- only test files that failed last test
--warn-long [timeout] -- warning if doctest is slow
--only-errors -- only output failures, not successes
--gc=GC -- control garbarge collection (ALWAYS:
collect garbage before every test; NEVER:
disable gc; DEFAULT: Python default)
--short[=secs] -- run as many doctests as possible in about 300
seconds (or the number of seconds given.) This runs
the tests for each module from the top of the file
and skips tests once it exceeds the budget
allocated for that file.
--help -- show all doctesting options
--tnew [...] -- equivalent to -t --new
-tp <N> [...] -- like -t above, but tests in parallel using
N threads, with 0 interpreted as min(8, cpu_count())
--testall [options] -- equivalent to -t --all
--coverage <files> -- give information about doctest coverage of files
--coverageall -- give summary info about doctest coverage of
all files in the Sage library
--startuptime [module] -- display how long each component of Sage takes to
start up; optionally specify a module to get more
details about that particular module
--tox [options] <files|dirs> -- general entry point for testing
and linting of the Sage library
-e <envlist> -- run specific test environments; default:
doctest,coverage,startuptime,pycodestyle-minimal,relint,codespell,rst
doctest -- run the Sage doctester
(same as "sage -t")
coverage -- give information about doctest coverage of files
(same as "sage --coverage[all]")
startuptime -- display how long each component of Sage takes to start up
(same as "sage --startuptime")
pycodestyle-minimal -- check against Sage's minimal style conventions
relint -- check whether some forbidden patterns appear
(includes all patchbot pattern-exclusion plugins)
codespell -- check for misspelled words in source code
rst -- validate Python docstrings markup as reStructuredText
coverage.py -- run the Sage doctester with Coverage.py
coverage.py-html -- run the Sage doctester with Coverage.py, generate HTML report
pyright -- run the static typing checker pyright
pycodestyle -- check against the Python style conventions of PEP8
cython-lint -- check Cython files for code style
-p auto -- run test environments in parallel
--help -- show tox help
--pytest [options] <files|dirs> -- run pytest on the Sage library
--help -- show pytest help
Some developer utilities:
--grep [options] <string>
-- regular expression search through the Sage
library for "string". Any options will
get passed to the "grep" command.
--grepdoc [options] <string>
-- regular expression search through the
Sage documentation for "string".
--search_src ... -- same as --grep
--search_doc ... -- same as --grepdoc
--fixdoctests file.py
-- Run doctests and replace output of failing doctests
with actual output.
--fiximports <files|dirs>
-- Replace imports from sage.PAC.KAGE.all by specific
imports when sage.PAC.KAGE is an implicit namespace
package
--fixdistributions <files|dirs>
-- Check or update '# sage_setup: distribution'
directives in source files
--sh [...] -- run a shell with Sage environment variables
as they are set in the runtime of Sage
--cleaner -- run the Sage cleaner. This cleans up after Sage,
removing temporary directories and spawned processes.
(This gets run by Sage automatically, so it is usually
not necessary to run it separately.)
File conversion:
--rst2ipynb [...] -- Generates Jupyter notebook (.ipynb) from standalone
reStructuredText source.
(not installed currently, run sage -i rst2ipynb)
--ipynb2rst [...] -- Generates a reStructuredText source file from
a Jupyter notebook (.ipynb).
--sws2rst <sws doc> -- Generates a reStructuredText source file from
a Sage worksheet (.sws) document.
(not installed currently, run sage -i sage_sws2rst)
Valgrind memory debugging:
--cachegrind -- run Sage using Valgrind's cachegrind tool. The log
files are named sage-cachegrind.PID can be found in
$DOT_SAGE
--callgrind -- run Sage using Valgrind's callgrind tool. The log
files are named sage-callgrind.PID can be found in
$DOT_SAGE
--massif -- run Sage using Valgrind's massif tool. The log
files are named sage-massif.PID can be found in
$DOT_SAGE
--memcheck -- run Sage using Valgrind's memcheck tool. The log
files are named sage-memcheck.PID can be found in
$DOT_SAGE
--omega -- run Sage using Valgrind's omega tool. The log
files are named sage-omega.PID can be found in
$DOT_SAGE
--valgrind -- this is an alias for --memcheck
Getting help:
-v, --version -- display Sage version information
--dumpversion -- print brief Sage version
-h, -?, --help -- print a short help message
--advanced -- list all command line options
Building the Sage library:
-b -- build Sage library -- do this if you have
modified any source code files in SAGE_ROOT/src/sage/
-ba -- same as -b, but rebuild *all* Cython
code.
-br -- build and run Sage
-bt [...] -- build Sage and test; same options as -t
-btp <N> [...] -- build Sage and test in parallel; same options as -tp
-btnew [...] -- build Sage and test modified files, as in -t --new
-bn [...], --build-and-notebook [...]
-- build the Sage library (as by running "sage -b")
and then start the notebook
Package handling:
--package [args] -- call the package manager with given arguments.
Run without arguments for help.
--experimental -- list all experimental packages that can be installed
-i [opts] [pkgs] -- install the given Sage packages. Options:
-c -- run the packages' test suites,
overriding the settings of
SAGE_CHECK and SAGE_CHECK_PACKAGES
-d -- only download, do not install packages
-f -- force build: install the packages even
if they are already installed
-s -- do not delete the temporary build directories
after a successful build
-y -- reply yes to prompts about experimental
and old-style packages; warning: there
is no guarantee that these packages will
build correctly; use at your own risk
-n -- reply no to prompts about experimental
and old-style packages
-f [opts] [pkgs] -- shortcut for -i -f: force build of the given Sage
packages
-p [opts] [packages]-- install the given Sage packages, without dependency
checking. Options are the same as for the -i command.
--optional -- list all optional packages that can be installed
--standard -- list all standard packages that can be installed
--installed -- list all installed packages
Making Sage distributions:
--sdist -- build a source distribution of Sage
Building the documentation:
--docbuild [lang/]<document> <html|pdf|...> -- Build the Sage documentation
Other developer tools:
--root -- print the Sage root directory
--git-branch -- print the current git branch
--buildsh [...] -- run a shell with Sage environment variables
as they are set while building Sage and its packages