Exception access violation

I am getting an intermittent access violation error running some code that does some processing on random spike trains. I put a link to the gist that contains the notebook and wrapper functions I wrote for formatting the PyPlot graphs. I don't think the error has anything to do with the specific spike time vector because I saved each vector before attempting to plot it. After Julia crashed, I can reload that vector and it plots fine. The access violation happens infrequently--maybe one out of every 30-50 times I run the function. [Win 7 x64 with PyPlot 1.4.3; PyCall 0.6.2 with the default Anaconda 2. x64 installation; Julia 0.3.3.]

Gist: https://gist.githubusercontent.com/anonymous/343fabf11334e7db1afa/

---

Please submit a bug report with steps to reproduce this fault, and any error mes
sages that follow (in their entirety). Thanks.
Exception: EXCEPTION_ACCESS_VIOLATION at 0x30f42d1d -- unknown function (ip: 821
308701)
unknown function (ip: 821308701)
unknown function (ip: 822083755)
.... [skipping more lines like this]
unknown function (ip: 821724369)
pycall at C:\Users\bens.julia\v0.3\PyCall\src\PyCall.jl:75
unknown function (ip: 1811228495)
unknown function (ip: 14081488)
unknown function (ip: 820977898)

Please submit a bug report with steps to reproduce this fault, and any error mes
sages that follow (in their entirety). Thanks.
Exception: EXCEPTION_ACCESS_VIOLATION at 0x30f42d1d -- unknown function (ip: 821
308701)
unknown function (ip: 820977898)

---

Julia Version 0.3.3
Commit b24213b\* (2014-11-23 20:19 UTC)
Platform Info:
System: Windows (x86_64-w64-mingw32)
CPU: Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz
WORD_SIZE: 64
BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Haswell)
LAPACK: libopenblas
LIBM: libopenlibm
LLVM: libLLVM-3.3

Start with the linked gist and the PyPlot/PyCall call site mentioned at PyCall\src\PyCall.jl:75, using the reported Windows x64, Julia 0.3.3, PyPlot 1.4.3, and Anaconda setup. Reproduce the intermittent access violation and capture the complete error output; done means identifying a reliable trigger and documenting the cause or a confirmed fix.

Do mô hình lập chỉ mục viết ra từ nội dung của issue.