来自boost-python子模块的python qt float精度

Question

I have made a cpp submodule with boost-python for my PyQt program that among others extracts some data from a zip data file.

It works fine when testing it in python:

import BPcmods
BPzip = BPcmods.BPzip()
BPzip.open("diagnostics/p25-dev.zip")
l=BPzip.getPfilenames()
t=BPzip.getTempArray([l[1][4],l[1][3]],40.)
print(t[11][:10])

>> [40.1, 40.44, 40.78, 41.11, 41.44, 41.77, 41.77, 42.09, 42.41, 42.73]

if I put the same code in the start of my main part, it also gives the same data, but if I put the code right after I call

main.app = QtGui.QApplication(sys.argv)

It suddenly yields:

[40.0, 40.0, 40.0, 41.0, 41.0, 41.0, 41.0, 42.0, 42.0, 42.0]

Why?? what happened? The numbers are still floats, but suddenly they have been rounded to nearest integer?!?

Answer 1

Well it was related to using std::stod to convert my strings of data from my datafile to doubles. I don't know why, but changing to:

boost::algorithm::trim(s);
double val = boost::lexical_cast<double>(s);

made it work as it was supposed to, also in pyqt.