doctest largest_connected_submatrix fails on py3.4

[marscher]

FAIL: Doctest: msmtools.estimation.api.largest_connected_submatrix

----------------------------------------------------------------------

Traceback (most recent call last):

  File "/home/travis/miniconda/envs/ci/lib/python3.4/doctest.py", line 2187, in runTest

    raise self.failureException(self.format_failure(new.getvalue()))

AssertionError: Failed doctest test for msmtools.estimation.api.largest_connected_submatrix

  File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 524, in largest_connected_submatrix

----------------------------------------------------------------------

File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 568, in msmtools.estimation.api.largest_connected_submatrix

Failed example:

    C_cc_directed

Expected:

    array([[10,  1],

           [ 2,  0]])

Got:

    array([[10,  1],

           [ 2,  0]], dtype=int64)

----------------------------------------------------------------------

File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 573, in msmtools.estimation.api.largest_connected_submatrix

Failed example:

    C_cc_undirected

Expected:

    array([[10,  1,  0],

           [ 2,  0,  3],

           [ 0,  0,  4]])

Got:

    array([[10,  1,  0],

           [ 2,  0,  3],

           [ 0,  0,  4]], dtype=int64)

[franknoe]

Has the default datatype when creating an array changed?

Am 28/07/15 um 12:32 schrieb Martin K. Scherer:

|FAIL: Doctest: msmtools.estimation.api.largest_connected_submatrix

---

Traceback (most recent call last):

File "/home/travis/miniconda/envs/ci/lib/python3.4/doctest.py", line 2187, in runTest

 raise self.failureException(self.format_failure(new.getvalue()))

AssertionError: Failed doctest test for msmtools.estimation.api.largest_connected_submatrix

File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 524, in largest_connected_submatrix

---

File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 568, in msmtools.estimation.api.largest_connected_submatrix

Failed example:

 C_cc_directed

Expected:

 array([[10,  1],

        [ 2,  0]])

Got:

 array([[10,  1],

        [ 2,  0]], dtype=int64)

---

File "/home/travis/build/markovmodel/msmtools/msmtools/estimation/api.py", line 573, in msmtools.estimation.api.largest_connected_submatrix

Failed example:

 C_cc_undirected

Expected:

 array([[10,  1,  0],

        [ 2,  0,  3],

        [ 0,  0,  4]])

Got:

 array([[10,  1,  0],

        [ 2,  0,  3],

        [ 0,  0,  4]], dtype=int64)

|
|

—
Reply to this email directly or view it on GitHub
#40.

---

Prof. Dr. Frank Noe
Head of Computational Molecular Biology group
Freie Universitaet Berlin

Phone: (+49) (0)30 838 75354
Web: research.franknoe.de

Mail: Arnimallee 6, 14195 Berlin, Germany

[marscher]

seems like that. So maybe we should pin it down to 32bits?

[franknoe]

Since this is just for the doctest, why not int?

Am 28/07/15 um 13:43 schrieb Martin K. Scherer:

seems like that. So maybe we should pin it down to 32bits?

—
Reply to this email directly or view it on GitHub
#40 (comment).

---

Prof. Dr. Frank Noe
Head of Computational Molecular Biology group
Freie Universitaet Berlin

Phone: (+49) (0)30 838 75354
Web: research.franknoe.de

Mail: Arnimallee 6, 14195 Berlin, Germany

[marscher]

So we always use int64 for dtrajs, but the difference with the doctest is, that on Python3.4 Numpy spits out the dtype, which is not the case for 2.7 Numpy.

[marscher]

will fix that.