@piti118 How about generating and hosting the iminuit docs on https://readthedocs.org/ ?

I've been doing this for years and it works really well.

Note that currently the docs hosted at http://iminuit.github.io/iminuit/ are outdated,
the first example on the front page still contains the typo m.migrads() (extra "s") which has been fixed over two years ago.
This is the main advantage of readthedocs ... automatic docs build / update on git push.

Other advantages:

• Automatic docs build / update on git push
• Nice layout (in my opinion)

I've set up readthedocs a few times and would be happy to do it for iminuit if you agree.

This package looks terrific! Bundling Minuit (and being on pypi) is a killer feature and I'm going to start using this in my package.

Would it be possible to add support for single-sided parameter limits? It looks like this could be done in Minuit._prepare_initial_param() by checking if self.initiallimit[vname][0] and self.initiallimit[vname][1] are None.

This is a bit related to #90 .

I propose to clean up the top-level namespace to not show the user things she only very rarely needs to access in her code.

At the moment we have this:

In [1]: import iminuit

In [2]: iminuit.
iminuit.ConsoleFrontend      iminuit.InitialParamWarning  iminuit.color                iminuit.info                 iminuit.warnings
iminuit.HesseFailedWarning   iminuit.Minuit               iminuit.describe             iminuit.latex                
iminuit.IMinuitWarning       iminuit.Struct               iminuit.iminuit_warnings     iminuit.util                 

In [3]: iminuit?
Type:       module
String Form:<module 'iminuit' from '/Users/deil/Library/Python/2.7/lib/python/site-packages/iminuit/__init__.pyc'>
File:       /Users/deil/Library/Python/2.7/lib/python/site-packages/iminuit/__init__.py
Docstring:  <no docstring>
In [4]: iminuit.util.
iminuit.util.StringIO                  iminuit.util.dock_if_bound             iminuit.util.fitarg_rename             iminuit.util.re
iminuit.util.Struct                    iminuit.util.extract_error             iminuit.util.inspect                   iminuit.util.remove_var
iminuit.util.arguments_from_docstring  iminuit.util.extract_fix               iminuit.util.is_bound                  iminuit.util.true_param
iminuit.util.better_arg_spec           iminuit.util.extract_iv                iminuit.util.make_func_code            
iminuit.util.describe                  iminuit.util.extract_limit             iminuit.util.param_name                

I propose the following changes:

• Add a docstring with the basic example from the README to iminuit? so that new user's know how to get started.
• Have all the warning classes in iminuit.warnings, i.e. don't import the following warnings in the top-level namespace: iminuit.InitialParamWarning, iminuit.HesseFailedWarning, iminuit.IMinuitWarning
• The stdlib warnings shouldn't be there
• ...

@piti118 Would it be OK if I give it a shot and remove the things I think are superfluous and make a PR?

Examples are here: http://nbviewer.ipython.org/4655035/

The issues are:

• Minuit.ncalls is updated to the number of function calls for MIGRAD and HESSE, it should also be updated for MINOS
• Minuit.print_fmin() displays Minuit.ncalls as NFCN and Minuit.get_num_call_fcn() as NCALLS. This is confusing that Minuit.ncalls doesn't correspond to NCALLS
• Minuit.ncalls doesn't have a docstring
• Minuit.ncalls is a float. Why not an int?
• Maybe Minuit.get_num_call_fcn() could be exposed as a property like Minuit.ncalls_total?

port it to long double instead of double for better precision

scikit-hep/probfit#13

When using the contours function with subtract_min=True and plotting using matplotlib's contours function, I get multiple lines that extend beyond the sigma (bound) of 1.0 that I thought I set.

Try this:

from iminuit import Minuit
def f(x,y,z):
    return (x-2)**2 + (y-3)**2 + (z-4)**2
m = Minuit(f, limit_x=(3, 2))
m.migrad()

I get

Exception TypeError: "'str' object is not callable" in 'iminuit._libiminuit.Minuit.initialParameterState' ignored

followed by a segfault (log).

If I replace the upper limit of 2 with None I get

Exception TypeError: 'a float is required' in 'iminuit._libiminuit.Minuit.initialParameterState' ignored

followed by a segfault (log)

...turns

def f(x,y,z) in to
    blah blah

in to this, at run time and compile it.

cdef double f(vector[double] v):
    cdef double x = v[0]
    cdef double y = v[1]
    cdef double z = v[2]
    blah blah

And use special PythonFCN that takes this function pointer that looks like this. This will get rid of all python function call.

@cdeil @mattbellis You might be interested in this package. No root required at all.

@piti118 Can we add an "iminuit design" page to the iminuit docs (and same for probfit) that in a few minutes explains the most important concepts?

I tried summarizing it in a Figure, but now I think a real simple code example that contains all important concepts would be better!?

Description:

• Very loose coupling:
  Minuit only cares about FitStatistic.call for fit_statistic function.
  How the FitStatistic is defined doesn't matter, there don't even have to be Data or Model
• Parameters defined by fit_statistic call parameters.
• Parameter initial values, initial step size, fix / free flag, min / max limits set in Minuit constructor
• Model and FitStatistic don't store any state. Fit results (including parameter values, symmetric errors, asymmetric errors) stored in Minuit as C Structs that behave like Python dicts.
• To evaluate the model use model(data, **minuit.values), to evaluate the fit statistic use fit_statistic(**minuit.values) and to change parameters and re-fit use

minuit.migrad()
minuit.fit_arg['p'] = 42
minuit.fit_arg['fix_p2'] = False
minuit.migrad()

• No choice of minimiser or error estimator. Only MIGRAD, HESSE, MINOS available.
  E.g. other minimizers are faster and can take analytical derivatives or more complex constraints than min / max ... I don't think these things are possible with MIGRAD, right?

Hi,

I am currently running iminuit 1.1.0 and when I try to set a limit for a parameter the lower limit is not set. A basic example is given in
slac.stanford.edu/~leddig/iMinuit Bug.ipynb
In the table, giving the fitted parameters the field "Limit-" is empty.
I already tried this smal program with iminuit 1.0.8 and here it works perfectly. I guess the changes for the one sided limit somehow had an unfortunate side-effect.

Cheers,
Torsten

Hello,
I am learning how to use iminuit (having used pyminuit before). I see iminuit prints some useful information - eg. when running from a terminal, if you don't set parameter uncertainties when declaring the Minuit object it warns you, when performing a minimization (.fit()), the results are printed etc.

I am wondering if it is possible to suppress all these messages? While this is useful for general fits, I also have to do many monte-carlo fits, where I loop over a set of data and perform ~100 or 200 fits in my monte-carlo estimate. In this case, I do not want to print these types of messages for all the fits.

Thanks,
André

I think I found two issues with the bound parameter:
http://nbviewer.ipython.org/4654173/

1. Calling m.draw_profile('x', bound=[2.5, 3]) if the x minimum is at 2 works, but gives this error:

ValueError: attempt to get argmin of an empty sequence

I guess it could be argued if that is a valid use case ...
2. Calling m.contour('x', 'y', bound=[[0, 4],[1, 5]]) doesn't work, although according to the docs:

If bound is 2x2 array [[v1min,v1max],[v2min,v2max]].

@cdeil since you are good with travis. Can we do that? I'm wondering if it compiles with pypy.

I'm quite surprise but some people are actually using it https://twiki.ph.rhul.ac.uk/twiki/bin/view/Public/LogBookAstronomyEntry0010

Here's an example that illustrates the issue: http://nbviewer.ipython.org/4653607/

After the MIGRAD call it is printed EDM = 2.41466832126e-09, which is below the GOAL EDM = 1e-05, i.e. the Above EDM status field should be False, but it says True.

Maybe the Above EDM check is not reliable or implemented properly for certain MIGRAD failure cases?

This is the end of what is printed on the console:

VariableMetricBuilder: warning: no improvement in line search  
VariableMetricBuilder: finishes without convergence.
VariableMetricBuilder: edm= 0.00560136 requested: 1e-05
Elapsed Time: 1 sec Call: 443/10000 (0.0023 sec/call).
Est. time to maxcall: 21.57 sec.
FCN=4.008063e+04 FROM MIGRAD  STATUS=INVALID   443 CALLS
EDM=2.414668e-09 ERROR MATRIX ACCURATE POSDEF
NO.       NAME         VALUE         ERROR  FIXED
  0       xpos  1.953868e-02  1.280456e-02       
  1       ypos  -3.615664e-02  1.278539e-02       
  2  amplitude  3.974392e+01  2.267837e-01       
  3 sigma_source  3.305788e-05  4.382916e-01       
  4 background_amplitude  2.499164e+01  2.750294e-02       
--------------------------------------

Just note here in case I forgot.

It's not possible to run tutorial.py because it uses the cythonmagic extension, which I think only works in the notebook? Maybe either remove the tutorial.py file, or delete the cythonmagic example?

$ python tutorial.py 
  File "tutorial.py", line 189
    %load_ext cythonmagic
    ^
SyntaxError: invalid syntax

I find that the IPython notebook is very nice for interactive data analysis, but not so much for writing 100+ line code, i.e. I wouldn't write the models, data wrangling and cost function for a complex fitting task as an IPython notebook. Could you add an example with Cython (and maybe numpy arrays for the data), i.e. how to write / build / run it?

Would it be better to bundle Minuit2 instead of SEAL Minuit?
http://seal.web.cern.ch/seal/work-packages/mathlibs/minuit/release/download.html

It seems that since the last release of SEAL Minuit in 2006 quite a few bugs in Minuit have been fixed:
http://seal.web.cern.ch/seal/work-packages/mathlibs/minuit/release/relnotes2.html

@piti118 Do you think the fact that Minuit2 has a bunch of bugs fixed and is actively maintained warrants the work of switching in iminuit? What about the API changes and license?

Hi,
it would be great to have the possibility of fixing parameters like

m=iminuit.Minuit(f)
m.fixed["x"] = True

a bit like pyminuit

Thanks for your great work,
marco

In some very hard cases, I found that a brute force grid search in pyminuit helped Migrad.

https://code.google.com/p/pyminuit/wiki/FunctionReference

Is there any work done to implement this in iminuit? If not, I could give it a try. Something like the scipy version, but with the same interface.

http://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.brute.html

At the moment Minuit.matrix() returns the covariance matrix and Minuit.print_matrix() prints the correlation matrix.
Can we make both them consistent and have the possibility to get and print both covariance and correlation matrix?
I think the best solution might actually be being very explicit here to avoid any confusion:

Minuit.covariance_matrix # property or method
Minuit.correlation_matrix # property or method
Minuit.print_covariance_matrix() # pretty-print
Minuit.print_correlation_matrix() # pretty-print

@piti118 What do you think?

Hi, I have tried to install pyminuit on Windows 7 (64 bit) with pip install iminuit but I get some errors. I'm using an Anaconda distribution of python.
it returns to me a lot of Warnings (deprecated conversion) and in the end a lot of undefined references like this one:
build\temp.win-amd64-2.7\Release\iminuit_libiminuit.o:_libiminuit.cpp:(.text+0x11749): undefined reference to `_imp__PyErr_Occurred'
what could I do?

easy fix too... let's note it here.

Hi,

it would be very useful if one could have a variable numbers of parameters in the minimize function. I could minimize some (nuisance) parameters analytically to speed up the minimization or decide to include them in the fit to properly determine the parameter uncertainties with all correlations considered. Would something like that make sense? If yes i could implement it.

def f(x, y, args):
chi2 = x__2+y*2
for arg in args:
chi2 *= arg
return chi2

pars = ('x', 'y', 'nuis0', 'nuis1')
m = Minuit(f, forced_parameters=pars)

This is a comment from @piti118 from the README:

HTML output looks ugly. Yah.. I'm lazy. Please help me change it.

The current one will confuses users. Note that contours is not exactly 1 sigma contour since other parameters are fixed.

MnContours should be added to cython code to return the actual contour.

@mattbellis noting the issue you told me here so I don't forget.

Would it be possible to optionally add a fitting progress display, i.e. info on how long MIGRAD or MINOS will still run?
I know MINUIT can report progress, but as far as I know it's either nothing or a lot. What I have in mind is one line, something like this:

MIGRAD: Function call XXX of YYY max (ZZZ % done, ETA 42 sec), fit statistic = AAA

I have an example where the migrad call crashes:
https://gist.github.com/4443071

Here's the traceback (full report at https://gist.github.com/4443089 ):

Thread 0 Crashed:: Dispatch queue: com.apple.main-thread
0   libsystem_c.dylib               0x00007fff8d304aa1 memmove$VARIANT$sse42 + 217
1   _libiminuit.so                  0x000000010329bdb3 MnUserParameterState::fix(unsigned int) + 211
2   _libiminuit.so                  0x0000000103260791 __pyx_f_7iminuit_11_libiminuit_6Minuit_initialParameterState(__pyx_obj_7iminuit_11_libiminuit_Minuit*) + 2705
3   _libiminuit.so                  0x0000000103269cc5 __pyx_pw_7iminuit_11_libiminuit_6Minuit_3migrad(_object*, _object*, _object*) + 1701
4   org.python.python               0x0000000102853a46 PyEval_EvalFrameEx + 10598
5   org.python.python               0x0000000102851096 PyEval_EvalCodeEx + 1990
6   org.python.python               0x00000001028508c6 PyEval_EvalCode + 54
7   org.python.python               0x0000000102877aee PyRun_FileExFlags + 174
8   org.python.python               0x0000000102877659 PyRun_SimpleFileExFlags + 777
9   org.python.python               0x000000010288b378 Py_Main + 2952
10  libdyld.dylib                   0x00007fff8e3ca7e1 start + 1

This is on Mac OS X with Python 2.7 with iminuit 17ad3ed

It would be nice to have (maybe optionally) some color in the iminuit terminal console output.

Upgrading from v1.0.7 to v1.0.9 caused fits that used to work to fail. I'm having trouble coming up with a simple test case to illustrate the problem, but one piece of mysterious behavior is that the initial parameter values change between the two versions in the case where there are limits. Consider the simple function:

def f(x,y,z):
    print "CALL", x, y, z
    return (x-2.)**2 + (y-3.)**2 + (z-4.)**2

m = Minuit(f, x=2., limit_x=(-1.e5, 1.e5))
m.migrad()

In v1.0.7, this results in:

**************************************************
*                     MIGRAD                     *
**************************************************

CALL 1.99999999999 0.0 0.0
CALL 2.00999999999 0.0 0.0
CALL 1.98999999999 0.0 0.0
CALL 2.00248975561 0.0 0.0
CALL 1.99751024437 0.0 0.0
CALL 1.99999999999 0.01 0.0
CALL 1.99999999999 -0.01 0.0
CALL 1.99999999999 0.00248975551057 0.0
CALL 1.99999999999 -0.00248975551057 0.0
CALL 1.99999999999 0.0 0.01
[...]

as you'd expect. However, in v1.0.9, this results in

**************************************************
*                     MIGRAD                     *
**************************************************

CALL 49999.5 0.0 0.0
CALL 49999.0 0.0 0.0
CALL 50000.0 0.0 0.0
CALL 49999.5 1.0 0.0
CALL 49999.5 -1.0 0.0
CALL 49999.5 0.0 1.0
CALL 49999.5 0.0 -1.0
======================================
CALL 2.00001501323 3.0 4.0
CALL 1.95001501322 3.0 4.0
CALL 2.05001501323 3.0 4.0
CALL 2.00001501323 3.1 4.0
CALL 2.00001501323 2.9 4.0
CALL 2.00001501323 3.01 4.0
CALL 2.00001501323 2.99 4.0
[...]

Initially, minuit uses a value of x=49999.5 even though x is explicitly set to 2!

Sorry this bug report is not more complete. I'll add more to it as I investigate further.

@piti118 Can you init Minuit so that this doesn't crash?

import iminuit
def f(x): return 2 + 3 * x
fitter = iminuit.Minuit(f)
fitter.print_fmin()

(I didn't try to the other print functions or other things before calling migrad, they might also crash if Minuit isn't initialised fully.)

making sure when people pass likelihood it's set to something smart

Easy fix.... but let's note it here.

Hi all

I am trying to fit some parameters with Migrad, although the results are valid Migrad say it did not converge and the Minos fails to run. The edm is quite large 4352222. The errdef is set to 0.5. Is this a bug. But it works well for other data. I also checked my data and its clean.

Any hints?

Thanks

Jothy

conda is very popular ... it would be nice if we could make it easy to install iminuit via conda install iminuit ... especially for Windows users (see #115).

I think the way to do it is to write a recipe and binstar can be used to build and distribute binary packages, but I've never done this.
Does anyone know / has time to do this?

There is https://binstar.org/pypi/iminuit, but I think that's just a copy of the source tarball from PyPI, i.e. not useful.

It looks like one important feature of Minuit that is not exposed by iminuit is the possibility for the user to compute gradients themselves (see http://root.cern.ch/root/html/ROOT__Minuit2__FCNGradientBase.html).

@piti118 Do you think it would be useful to add this? Any thoughts on how to expose this to the end user?

I just installed the current version, and when I attempt to load the module I get the following error:
~> python
Python 2.6.6 (r266:84292, Dec 7 2011, 20:48:22)
[GCC 4.4.6 20110731 (Red Hat 4.4.6-3)] on linux2
Type "help", "copyright", "credits" or "license" for more information.

import iminuit
Traceback (most recent call last):
File "", line 1, in
File "iminuit/init.py", line 8, in
from _libiminuit import *
ImportError: No module named _libiminuit

when I look in the install location I don't see a _libiminuit module, just a .so file
~> ll /usr/lib64/python2.6/site-packages/iminuit/
total 8740
-rw-r--r-- 1 root root 724 Jul 14 14:51 color.py
-rw-r--r-- 1 root root 1391 Jul 14 14:53 color.pyc
-rw-r--r-- 1 root root 6130 Jul 14 14:51 ConsoleFrontend.py
-rw-r--r-- 1 root root 6435 Jul 14 14:53 ConsoleFrontend.pyc
-rw-r--r-- 1 root root 11232 Jul 14 14:51 HtmlFrontend.py
-rw-r--r-- 1 root root 13357 Jul 14 14:53 HtmlFrontend.pyc
-rw-r--r-- 1 root root 240 Jul 14 14:51 iminuit_warnings.py
-rw-r--r-- 1 root root 885 Jul 14 14:53 iminuit_warnings.pyc
-rw-r--r-- 1 root root 22 Jul 14 14:51 info.py
-rw-r--r-- 1 root root 169 Jul 14 14:53 info.pyc
-rw-r--r-- 1 root root 223 Jul 14 15:03 init.py
-rw-r--r-- 1 root root 434 Jul 14 14:53 init.pyc
-rw-r--r-- 1 root root 6564 Jul 14 14:51 latex.py
-rw-r--r-- 1 root root 7103 Jul 14 14:53 latex.pyc
-rwxr-xr-x 1 root root 8799070 Jul 14 14:53 _libiminuit.so
-rw-r--r-- 1 root root 4485 Jul 14 14:51 _plotting.py
-rw-r--r-- 1 root root 4744 Jul 14 14:53 _plotting.pyc
-rw-r--r-- 1 root root 6789 Jul 14 14:51 util.py
-rw-r--r-- 1 root root 9250 Jul 14 14:53 util.pyc

I pulled the files via git:
~> git clone https://github.com/iminuit/iminuit.git
Initialized empty Git repository in /tmp/iminuit/iminuit/.git/
remote: Reusing existing pack: 1784, done.
remote: Total 1784 (delta 0), reused 0 (delta 0)
Receiving objects: 100% (1784/1784), 2.86 MiB, done.
Resolving deltas: 100% (856/856), done.

Then installed:
~> sudo python setup.py install
running install
running build
running build_py
creating build
creating build/lib.linux-x86_64-2.6
creating build/lib.linux-x86_64-2.6/iminuit
copying iminuit/iminuit_warnings.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/init.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/ConsoleFrontend.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/info.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/color.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/HtmlFrontend.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/latex.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/util.py -> build/lib.linux-x86_64-2.6/iminuit
copying iminuit/_plotting.py -> build/lib.linux-x86_64-2.6/iminuit
running build_ext
building 'iminuit._libiminuit' extension
creating build/temp.linux-x86_64-2.6
creating build/temp.linux-x86_64-2.6/iminuit
creating build/temp.linux-x86_64-2.6/Minuit
creating build/temp.linux-x86_64-2.6/Minuit/src
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c iminuit/_libiminuit.cpp -o build/temp.linux-x86_64-2.6/iminuit/_libiminuit.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from iminuit/_libiminuit.cpp:360:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
In file included from iminuit/_libiminuit.cpp:364:
iminuit/PythonFCN.h: In member function 'std::string PythonFCN::errormsg(const std::vector<double, std::allocator >&) const':
iminuit/PythonFCN.h:139: warning: comparison between signed and unsigned integer expressions
iminuit/PythonFCN.h:142: warning: comparison between signed and unsigned integer expressions
iminuit/PythonFCN.h: In member function 'PyObject* PythonFCN::vector2tuple(const std::vector<double, std::allocator >&) const':
iminuit/PythonFCN.h:152: warning: comparison between signed and unsigned integer expressions
iminuit/_libiminuit.cpp: In function 'PyObject* __pyx_pf_7iminuit_11_libiminuit_6Minuit_18_prepare_param(_pyx_obj_7iminuit_11_libiminuit_Minuit)':
iminuit/_libiminuit.cpp:7881: warning: comparison between signed and unsigned integer expressions
iminuit/libiminuit.cpp: In function 'PyObject _pyx_pf_7iminuit_11_libiminuit_6Minuit_44get_param_states(pyx_obj_7iminuit_11_libiminuit_Minuit)':
iminuit/libiminuit.cpp:9644: warning: comparison between signed and unsigned integer expressions
iminuit/libiminuit.cpp: At global scope:
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
iminuit/libiminuit.cpp:22911: warning: deprecated conversion from string constant to 'char'
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaOuterProduct.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaOuterProduct.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/DavidonErrorUpdator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/DavidonErrorUpdator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnParameterScan.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnParameterScan.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/Numerical2PGradientCalculator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/Numerical2PGradientCalculator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnContours.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnContours.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/MnContours.cpp:5:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnvert.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnvert.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnLineSearch.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnLineSearch.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliBuilder.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliBuilder.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/Minuit/FumiliBuilder.h:8,
from Minuit/src/FumiliBuilder.cpp:1:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SimplexParameters.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SimplexParameters.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SqrtLowParameterTransformation.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SqrtLowParameterTransformation.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/ParametricFunction.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/ParametricFunction.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/VariableMetricEDMEstimator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/VariableMetricEDMEstimator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnCovarianceSqueeze.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnCovarianceSqueeze.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaEigenValues.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaEigenValues.o
Minuit/src/LaEigenValues.cpp: In function 'LAVector eigenvalues(const LASymMatrix&)':
Minuit/src/LaEigenValues.cpp:19: warning: unused variable 'info'
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnPrint.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnPrint.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/MnPrint.cpp:4:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnUserTransformation.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnUserTransformation.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnPosDef.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnPosDef.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mndspr.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mndspr.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnFumiliMinimize.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnFumiliMinimize.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/Minuit/FumiliBuilder.h:8,
from Minuit/Minuit/FumiliMinimizer.h:6,
from Minuit/Minuit/MnFumiliMinimize.h:5,
from Minuit/src/MnFumiliMinimize.cpp:1:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/CombinedMinimumBuilder.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/CombinedMinimumBuilder.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/CombinedMinimumBuilder.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnTiny.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnTiny.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SimplexSeedGenerator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SimplexSeedGenerator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliMinimizer.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliMinimizer.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/Minuit/FumiliBuilder.h:8,
from Minuit/Minuit/FumiliMinimizer.h:6,
from Minuit/src/FumiliMinimizer.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnPlot.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnPlot.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnlsame.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnlsame.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/ModularFunctionMinimizer.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/ModularFunctionMinimizer.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/ModularFunctionMinimizer.cpp:7:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mndspmv.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mndspmv.o
Minuit/src/mndspmv.cpp: In function 'int mndspmv(const char, unsigned int, double, const double, const double, int, double, double, int)':
Minuit/src/mndspmv.cpp:145: warning: suggest parentheses around '&&' within '||'
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnGlobalCorrelationCoeff.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnGlobalCorrelationCoeff.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnHesse.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnHesse.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliStandardChi2FCN.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliStandardChi2FCN.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnUserFcn.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnUserFcn.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/InitialGradientCalculator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/InitialGradientCalculator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnMinos.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnMinos.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/MnMinos.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnSeedGenerator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnSeedGenerator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SinParameterTransformation.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SinParameterTransformation.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaInverse.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaInverse.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaSumOfElements.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaSumOfElements.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mndasum.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mndasum.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mndaxpy.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mndaxpy.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnFcn.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnFcn.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnMachinePrecision.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnMachinePrecision.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliGradientCalculator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliGradientCalculator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnddot.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnddot.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/ScanBuilder.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/ScanBuilder.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/ScanBuilder.cpp:3:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnUserParameterState.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnUserParameterState.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mndscal.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mndscal.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/HessianGradientCalculator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/HessianGradientCalculator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnFunctionCross.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnFunctionCross.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/MnFunctionCross.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnApplication.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnApplication.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/MnApplication.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnEigen.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnEigen.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliErrorUpdator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliErrorUpdator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaVtMVSimilarity.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaVtMVSimilarity.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnxerbla.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnxerbla.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/VariableMetricBuilder.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/VariableMetricBuilder.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/VariableMetricBuilder.cpp:6:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SqrtUpParameterTransformation.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SqrtUpParameterTransformation.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mntplot.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mntplot.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/NegativeG2LineSearch.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/NegativeG2LineSearch.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/BasicMinimumError.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/BasicMinimumError.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/FumiliStandardMaximumLikelihoodFCN.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/FumiliStandardMaximumLikelihoodFCN.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnParabolaFactory.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnParabolaFactory.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnteigen.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnteigen.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/AnalyticalGradientCalculator.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/AnalyticalGradientCalculator.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/LaInnerProduct.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/LaInnerProduct.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnScan.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnScan.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnStrategy.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnStrategy.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/mnbins.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/mnbins.o
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/SimplexBuilder.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/SimplexBuilder.o
In file included from Minuit/Minuit/FunctionMinimum.h:4,
from Minuit/src/SimplexBuilder.cpp:2:
Minuit/Minuit/BasicFunctionMinimum.h: In member function 'void BasicFunctionMinimum::print(bool)':
Minuit/Minuit/BasicFunctionMinimum.h:169: warning: comparison between signed and unsigned integer expressions
gcc -pthread -fno-strict-aliasing -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -DNDEBUG -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -D_GNU_SOURCE -fPIC -fwrapv -fPIC -IMinuit -I/usr/include/python2.6 -c Minuit/src/MnUserParameters.cpp -o build/temp.linux-x86_64-2.6/Minuit/src/MnUserParameters.o
g++ -pthread -shared build/temp.linux-x86_64-2.6/iminuit/_libiminuit.o build/temp.linux-x86_64-2.6/Minuit/src/LaOuterProduct.o build/temp.linux-x86_64-2.6/Minuit/src/DavidonErrorUpdator.o build/temp.linux-x86_64-2.6/Minuit/src/MnParameterScan.o build/temp.linux-x86_64-2.6/Minuit/src/Numerical2PGradientCalculator.o build/temp.linux-x86_64-2.6/Minuit/src/MnContours.o build/temp.linux-x86_64-2.6/Minuit/src/mnvert.o build/temp.linux-x86_64-2.6/Minuit/src/MnLineSearch.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliBuilder.o build/temp.linux-x86_64-2.6/Minuit/src/SimplexParameters.o build/temp.linux-x86_64-2.6/Minuit/src/SqrtLowParameterTransformation.o build/temp.linux-x86_64-2.6/Minuit/src/ParametricFunction.o build/temp.linux-x86_64-2.6/Minuit/src/VariableMetricEDMEstimator.o build/temp.linux-x86_64-2.6/Minuit/src/MnCovarianceSqueeze.o build/temp.linux-x86_64-2.6/Minuit/src/LaEigenValues.o build/temp.linux-x86_64-2.6/Minuit/src/MnPrint.o build/temp.linux-x86_64-2.6/Minuit/src/MnUserTransformation.o build/temp.linux-x86_64-2.6/Minuit/src/MnPosDef.o build/temp.linux-x86_64-2.6/Minuit/src/mndspr.o build/temp.linux-x86_64-2.6/Minuit/src/MnFumiliMinimize.o build/temp.linux-x86_64-2.6/Minuit/src/CombinedMinimumBuilder.o build/temp.linux-x86_64-2.6/Minuit/src/MnTiny.o build/temp.linux-x86_64-2.6/Minuit/src/SimplexSeedGenerator.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliMinimizer.o build/temp.linux-x86_64-2.6/Minuit/src/MnPlot.o build/temp.linux-x86_64-2.6/Minuit/src/mnlsame.o build/temp.linux-x86_64-2.6/Minuit/src/ModularFunctionMinimizer.o build/temp.linux-x86_64-2.6/Minuit/src/mndspmv.o build/temp.linux-x86_64-2.6/Minuit/src/MnGlobalCorrelationCoeff.o build/temp.linux-x86_64-2.6/Minuit/src/MnHesse.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliStandardChi2FCN.o build/temp.linux-x86_64-2.6/Minuit/src/MnUserFcn.o build/temp.linux-x86_64-2.6/Minuit/src/InitialGradientCalculator.o build/temp.linux-x86_64-2.6/Minuit/src/MnMinos.o build/temp.linux-x86_64-2.6/Minuit/src/MnSeedGenerator.o build/temp.linux-x86_64-2.6/Minuit/src/SinParameterTransformation.o build/temp.linux-x86_64-2.6/Minuit/src/LaInverse.o build/temp.linux-x86_64-2.6/Minuit/src/LaSumOfElements.o build/temp.linux-x86_64-2.6/Minuit/src/mndasum.o build/temp.linux-x86_64-2.6/Minuit/src/mndaxpy.o build/temp.linux-x86_64-2.6/Minuit/src/MnFcn.o build/temp.linux-x86_64-2.6/Minuit/src/MnMachinePrecision.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliGradientCalculator.o build/temp.linux-x86_64-2.6/Minuit/src/mnddot.o build/temp.linux-x86_64-2.6/Minuit/src/ScanBuilder.o build/temp.linux-x86_64-2.6/Minuit/src/MnUserParameterState.o build/temp.linux-x86_64-2.6/Minuit/src/mndscal.o build/temp.linux-x86_64-2.6/Minuit/src/HessianGradientCalculator.o build/temp.linux-x86_64-2.6/Minuit/src/MnFunctionCross.o build/temp.linux-x86_64-2.6/Minuit/src/MnApplication.o build/temp.linux-x86_64-2.6/Minuit/src/MnEigen.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliErrorUpdator.o build/temp.linux-x86_64-2.6/Minuit/src/LaVtMVSimilarity.o build/temp.linux-x86_64-2.6/Minuit/src/mnxerbla.o build/temp.linux-x86_64-2.6/Minuit/src/VariableMetricBuilder.o build/temp.linux-x86_64-2.6/Minuit/src/SqrtUpParameterTransformation.o build/temp.linux-x86_64-2.6/Minuit/src/mntplot.o build/temp.linux-x86_64-2.6/Minuit/src/NegativeG2LineSearch.o build/temp.linux-x86_64-2.6/Minuit/src/BasicMinimumError.o build/temp.linux-x86_64-2.6/Minuit/src/FumiliStandardMaximumLikelihoodFCN.o build/temp.linux-x86_64-2.6/Minuit/src/MnParabolaFactory.o build/temp.linux-x86_64-2.6/Minuit/src/mnteigen.o build/temp.linux-x86_64-2.6/Minuit/src/AnalyticalGradientCalculator.o build/temp.linux-x86_64-2.6/Minuit/src/LaInnerProduct.o build/temp.linux-x86_64-2.6/Minuit/src/MnScan.o build/temp.linux-x86_64-2.6/Minuit/src/MnStrategy.o build/temp.linux-x86_64-2.6/Minuit/src/mnbins.o build/temp.linux-x86_64-2.6/Minuit/src/SimplexBuilder.o build/temp.linux-x86_64-2.6/Minuit/src/MnUserParameters.o -L/usr/lib64 -lpython2.6 -o build/lib.linux-x86_64-2.6/iminuit/_libiminuit.so
running install_lib
copying build/lib.linux-x86_64-2.6/iminuit/iminuit_warnings.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/_libiminuit.so -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/init.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/ConsoleFrontend.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/info.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/color.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/HtmlFrontend.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/latex.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/util.py -> /usr/lib64/python2.6/site-packages/iminuit
copying build/lib.linux-x86_64-2.6/iminuit/_plotting.py -> /usr/lib64/python2.6/site-packages/iminuit
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/iminuit_warnings.py to iminuit_warnings.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/init.py to init.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/ConsoleFrontend.py to ConsoleFrontend.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/info.py to info.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/color.py to color.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/HtmlFrontend.py to HtmlFrontend.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/latex.py to latex.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/util.py to util.pyc
byte-compiling /usr/lib64/python2.6/site-packages/iminuit/_plotting.py to _plotting.pyc
running install_egg_info
Removing /usr/lib64/python2.6/site-packages/iminuit-1.1.1-py2.6.egg-info
Writing /usr/lib64/python2.6/site-packages/iminuit-1.1.1-py2.6.egg-info

The subtract_min option works in draw_profile, but in draw_mnprofile it has no effect.

Here's an example:
http://nbviewer.ipython.org/4648067/

Dear all!

unfortunately iminuit is not Python3 compatible.
Is it planned to overcome this little problem in the future?

Cheers,
iRaffnix

@piti118 In the tutorial here
you say this:

You might be worried about that forking process will copy your data. Most modern OS use Copy On Write(look at wiki) what this means is that when it forks a process it doesn't copy memory there unless you are writing it

and in the code:

Copy on write ensures that data points will never be copied (unless your write to it)

Are you sure this is true?
I haven't tested myself, but from what I read this copy on write trick doesn't work in Python: see e.g. http://stackoverflow.com/a/660026/498873 .

Also the multiprocessing example references an IPython issue that has been closed.

@piti118 Can you update this example?

Hello Piti and all,

Based upon my experience with pyminuit, I am finding unexpectedly unstable behavior trying to minimize a non-linear function, a exp(-m * t). For example, one time I run the fit, and it gets the "correct" answer, another I run the fit, and it fails with RuntimeWarning: overflow errors where it seems the fit has wondered far away from where it should. I am attaching a simple script and data file which reproduces the error. It requires,

numpy, scipy, iminuit, pytables

I can make the input data file, 'test_data.h5' into ascii format if you don't have pytables. I am also posting the output of two successive running of the file in an ipython session.

I place the initial values near the minimum and also set the parameter errors close to the uncertainties I expect. I am not setting any limits, etc. Based on my experience with pyminuit, Minuit should be able to handle this minimization very easily. But for some reason, it is puking about half the time.

scratch - I don't see how to attach the files - so I am posting them at this website

http://cyclades.physics.wm.edu/~walkloud/iminuit_test/

Thanks,

Andre

In [15]: %run  test_exp.py
before setting parameter values, they are all 0.0
parameters =  ['m', 'a']
values =  {'a': 0.0, 'm': 0.0}
then we can give them initial values and errors
values =  {'a': -6.0000000000000002e-06, 'm': 0.28999999999999998}
errors =  {'a': 5.9999999999999997e-07, 'm': 0.0030000000000000001}
then we can call migrad
======================================
Elapsed Time: 3 sec Call: 424/10000 (0.0071 sec/call).
Est. time to maxcall: 67.75 sec.
FCN=5.484253e+00 FROM MIGRAD  STATUS=VALID   424 CALLS
EDM=4.809132e-06 ERROR MATRIX ACCURATE POSDEF
NO.       NAME         VALUE         ERROR  FIXED
  0          m  2.915365e-01  3.593883e-03       
  1          a  -5.815086e-06  8.896744e-07       
--------------------------------------

In [16]: %run  test_exp.py
before setting parameter values, they are all 0.0
parameters =  ['m', 'a']
values =  {'a': 0.0, 'm': 0.0}
then we can give them initial values and errors
values =  {'a': -6.0000000000000002e-06, 'm': 0.28999999999999998}
errors =  {'a': 5.9999999999999997e-07, 'm': 0.0030000000000000001}
then we can call migrad
======================================
test_exp.py:11: RuntimeWarning: overflow encountered in exp
  f = lambda t,m,a: a * np.exp(-t * m)
test_exp.py:43: RuntimeWarning: fcn returns Nan
fcn is called with following arguments:
    m = -18.120111
    a = -0.000000

test_exp.py:43: RuntimeWarning: fcn returns Nan
fcn is called with following arguments:
    m = nan
    a = -0.000000

Elapsed Time: 0 sec Call: 39/10000 (0.0000 sec/call).
Est. time to maxcall: 0.00 sec.
FCN=1.369120e+03 FROM MIGRAD  STATUS=VALID   39 CALLS
EDM=       nan ERROR MATRIX NOT ACCURATE POSDEF
NO.       NAME         VALUE         ERROR  FIXED
  0          m  0.000000e+00           nan       
  1          a  -6.269743e-14           nan       
--------------------------------------

iminuit no longer works on Mac (at least with default settings), because the default clang compiler doesn't support OpenMP and thus the linker can't create _libiminuit.so:

$ python setup.py install --user
...
ld: library not found for -lgomp
clang: error: linker command failed with exit code 1 (use -v to see invocation)
error: command '/usr/bin/clang++' failed with exit status 1
...

Can you make OpenMP use optional?
Can you split the part of the advanced tutorial that uses OpenMP into a separate tutorial (which would not work for clang users)?

@mattbellis Here you go

In some cases, migrad spits a lot of output complaining about the fit not being correct, even when print_level is set to 0. This is specially annoying when there are many fits in a program, as the actual informative output will be buried in the noise. I think the correct way to do this is through the Minuit class, so the code can access it, and printed if print_level is set appropriately.

The output stream of compiled code cannot be easily redirected, so for the moment I have commented out all the (relevant) cout lines in the Minuit/src folder, and now it works quietly. Perhaps, the ideal behaviour would be to capture that stream and redirect it to stdout if the print level is set to debugging, or /dev/null otherwise; but I don't know how to do it. I can submit a PR with my work so far if you want.

I have a target function which involves numerical integrals that are quite expensive to get a high precision. The current precision goes to 0.01 as a chi-square, which should be sufficient to optimize and to define error bounds. However, the fine-grid behaviour is quite noisy, and iminuit can barely converge. Even it does, the parameter errors returned is unrealistically small. I notice that in the fortran MINUIT there is [SET EPSmachine] to handle floating point precision. I cannot find a corresponding functionality in iminuit. How to achieve this with the current iminuit? Thanks a lot!

def f(x, y):
    return x**6 + y ** 4 + 5 * x**2 * y**3
m = iminuit.Minuit(f, x=1, y=1)
m.migrad()
print m1.matrix()

Returns the covariance matrix:

((14167.000515996086, -1696.9291971259613),
 (-1696.9291971259613, 712.8246991937284))

That is obviously wrong. Running minos or hesse don't seem to improve the values.

iminuit.describe works for the Python built-in pow function in Python 2.7:

In [6]: iminuit.describe(pow)
Out[6]: ['x', 'y', 'z']

But in Python 2.6 I get this error:

In [2]: iminuit.describe(pow)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-2-db36f21dd787> in <module>()
----> 1 iminuit.describe(pow)

/Users/deil/.local/lib/python2.6/site-packages/iminuit/util.pyc in describe(f, quiet)
     31 def describe(f,quiet=True):
     32     if not quiet: print better_arg_spec(f)
---> 33     return better_arg_spec(f)
     34 
     35 

/Users/deil/.local/lib/python2.6/site-packages/iminuit/util.pyc in better_arg_spec(f)
     26         #print e
     27         pass
---> 28     return inspect.getargspec(f)[0]
     29 
     30 

/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/inspect.pyc in getargspec(func)
    801         func = func.im_func
    802     if not isfunction(func):
--> 803         raise TypeError('arg is not a Python function')
    804     args, varargs, varkw = getargs(func.func_code)
    805     return ArgSpec(args, varargs, varkw, func.func_defaults)

TypeError: arg is not a Python function

@piti118 Is this supposed to work? Can you make it work?

I think the following would be nice for new users:

• Mention what the relation is to PyMinuit (what is different / better) and RTMinuit (or will that go away?)
• In the README you call it SEAL-Minuit, in the http://iminuit.github.com/iminuit/ top-level page lcg-minuit. Can you be consistent to avoid confusion?
• Some important things don't have a docstring: iminuit package, iminuit.describe and iminuit.Struct.
• The structs play a big role in the [http://iminuit.github.com/iminuit/api.html#return-value-struct](API docs). Could you explain what they are and in what respect they are / are not like Python dicts?
• [email protected] mailing list for asking questions?