JuliaPy/Pandas.jl

[Win7 / 64 bits] RuntimeError: attempt to load C runtime library incorrectly.

Ismael-VC opened this issue · 2 comments

pandas-error

This happens always on Windows 7 64 bits, yet everything else seem to be working fine until now.

C:\Users\CSM\MCEGOLD>julia
               _
   _       _ _(_)_     |  A fresh approach to technical computing
  (_)     | (_) (_)    |  Documentation: http://docs.julialang.org
   _ _   _| |_  __ _   |  Type "help()" for help.
  | | | | | | |/ _` |  |
  | | |_| | | | (_| |  |  Version 0.3.5 (2015-01-08 22:33 UTC)
 _/ |\__'_|_|_|\__'_|  |  Official http://julialang.org/ release
|__/                   |  x86_64-w64-mingw32

julia> using Pandas    # This is when the error message appears.
INFO: Loading help data...

julia> read_excel("data.xls")    # This still works though.
WARNING *** file size (6367) not 512 + multiple of sector size (512)
   NamePlate Information     Unnamed: 1
0                   Type        AC Mtr.
1                Frame #           127T
2                NP Freq             60
3                   Volt            440
4                    FLA             80
5                     KW           74.6
6                     HP            100
7                     PF            NaN
8                    Eff            NaN
9               NP Speed           1758
10              OP Speed           1758
11                 Slots            NaN
12                  Bars            NaN
13                   NaN            NaN
14                   NaN            NaN
15             Test Date      2/25/2015
16             Test Time    12:17:05 PM
17         Test Location    Motor Leads
18                  User  Administrator
19         Tester Serial            NaN
20               MTAP ID            NaN
21                             Baseline
22             Frequency           1200
23           Charge Time            600
24               Voltage            500
25            Motor Temp             18
26                                  NaN
27         Measured Mohm         > 300G
28        Corrected Mohm         > 300G
29     pF Ph 1 to Ground         < 1000
30         ohm Ph 1 to 2        0.00262
31         ohm Ph 2 to 3        0.00278
32         ohm Ph 3 to 1        0.00463
33          mH Ph 1 to 2          0.255
34          mH Ph 2 to 3          0.185
35          mH Ph 3 to 1           0.23
36    Average Inductance          0.225
37      % Res. Imbalance          38.48
38      % Ind. Imbalance          17.16

While using:

Python 2.7.8 |Anaconda 2.1.0 (64-bit)| (default, Jul  2 2014, 15:12:11) [MSC v.1500 64 bit (AMD64)] on win32

Unfortunately I'm not equipped to debug Windows-specific problems. Would be happy to accept a PR though.

This is probably a case of Julia trying to free memory that was allocated by Python, which doesn't go so well because they're using different C runtimes. It might be possible to build Julia using the custom mingw-static toolchain that the numpy developers have started using recently on Windows, to match what Python uses. For the official binaries it's more important to match what the openSUSE build service uses for cross-compiling package libraries, but it's worth trying to build a copy of Julia locally with the nonstandard mingw-static and see if things work.