This is a pretty printer for eigen types (with some support for stan-math types). Much of the logic comes from upstream eigen. Formatting is handled by numpy.
Your debugging output shouldn't look like this:
(gdb) p mat $1 = {<Eigen::PlainObjectBase<Eigen::Matrix<double, 4, 4, 0, 4, 4> >> = {<Eigen::MatrixBase<Eigen::Matrix<double, 4, 4, 0, 4, 4> >> = {<Eigen::DenseBase<Eigen::Matrix<double, 4, 4, 0, 4, 4> >> = {<Eigen::DenseCoeffsBase<Eigen::Matrix<double, 4, 4, 0, 4, 4>, 3>> = {<Eigen::DenseCoeffsBase<Eigen::Matrix<double, 4, 4, 0, 4, 4>, 1>> = {<Eigen::DenseCoeffsBase<Eigen::Matrix<double, 4, 4, 0, 4, 4>, 0>> = {<Eigen::EigenBase<Eigen::Matrix<double, 4, 4, 0, 4, 4> >> = {<No data fields>}, <No data fields>}, <No data fields>}, <No data fields>}, <No data fields>}, <No data fields>}, m_storage = {m_data = {array = { 0.68037543430941905, -0.21123414636181392, 0.56619844751721171, 0.59688006695214657, 0.82329471587356862, -0.60489726141323208, -0.32955448857022196, 0.53645918962380801, -0.44445057839362445, 0.10793991159086103, -0.045205896275679502, 0.25774184952384882, -0.27043105441631332, 0.026801820391231024, 0.90445945034942565, 0.8323901360074013}}}}, <No data fields>}
But rather like this!
(gdb) p mat $1 = Eigen::Matrix<double,4,4,ColMajor> (data ptr: 0x7fffffffdde0) [[ 0.68037543 0.82329472 -0.44445058 -0.27043105] [-0.21123415 -0.60489726 0.10793991 0.02680182] [ 0.56619845 -0.32955449 -0.0452059 0.90445945] [ 0.59688007 0.53645919 0.25774185 0.83239014]]
It is important to use the python/pip version which corresponds to your GDB
installation. You can find out more information using the python
command in
GDB. For example, from GDB repl, you can find where GDB python will search for
packages.
(gdb) python
>import sys
>print(sys.path)
>end
[..., '/path/to/site-packages', ...]
In Linux, you can also find the python used by checking the libraries linked to GDB.
$ ldd $(which gdb) | grep python
Once you find the python interpreter used by GDB, install eigengdb
using a corresponding python/pip (usually system pip).
git clone https://github.com/dmillard/eigengdb
cd eigengdb
python setup.py install # Make sure to use system python which matches GDB
python bin/eigengdb_register_printers
There is an example program you can play with in the examples/
directory.
cd examples
make
make debug
In the resulting gdb prompt that shows up, add your breakpoint and run the file
b 7
run # run until breakpoint
p mat # shows eigengdb formatting
Since NumPy is used to render the matrix, you can adjust NumPy's printing options in your gdb session to tweak the appearance, for example:
(gdb) p lin_invar_pref
$11 = Eigen::Matrix<double,4,12,ColMajor> (data ptr: 0x613000001380)
[[ 1.00000000e+00 3.16292670e-01 1.05028445e+00 5.02844469e-02
1.31629267e+00 -2.19900950e-01 -1.69616503e-01 -1.19332056e-01
-6.90476092e-02 -1.87631623e-02 3.15212845e-02 1.92783441e-01]
[ 0.00000000e+00 1.00000000e+00 1.88824955e-01 1.88824955e-01
1.00000000e+00 -8.36652088e-01 -6.47827133e-01 -4.59002178e-01
-2.70177223e-01 -8.13522683e-02 1.07472687e-01 3.26695824e-01]
[ 0.00000000e+00 0.00000000e+00 1.00000000e+00 1.00000000e+00
6.83197897e-16 -2.18264785e-01 7.81735215e-01 1.78173522e+00
2.78173522e+00 3.78173522e+00 4.78173522e+00 -4.36529569e-01]
[ 0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00
0.00000000e+00 3.97719125e-01 3.97719125e-01 3.97719125e-01
3.97719125e-01 3.97719125e-01 3.97719125e-01 7.95438250e-01]]
(gdb) python
>import numpy as np
>np.set_printoptions(linewidth=200, formatter={'float': lambda x: "{:5.2f}".format(x) if x !=0 else " "})
>end
(gdb) p lin_invar_pref
$12 = Eigen::Matrix<double,4,12,ColMajor> (data ptr: 0x613000001380)
[[ 1.00 0.32 1.05 0.05 1.32 -0.22 -0.17 -0.12 -0.07 -0.02 0.03 0.19]
[ 1.00 0.19 0.19 1.00 -0.84 -0.65 -0.46 -0.27 -0.08 0.11 0.33]
[ 1.00 1.00 0.00 -0.22 0.78 1.78 2.78 3.78 4.78 -0.44]
[ 0.40 0.40 0.40 0.40 0.40 0.40 0.80]]
eigengdb is licensed under MPL2.0.