-
This error occurs when running multiple concurrent PyEMMA runs (which do something completely harmless have no I/O, like solving a generalized eigenvalue problem). Looks like a failure to concurrently…
-
[`ss2zpk` is implemented as](https://github.com/scipy/scipy/blob/v0.17.0/scipy/signal/ltisys.py#L352)
```
return tf2zpk(*ss2tf(A, B, C, D, input=input))
```
Please don't go through transfer-function…
-
In `lapack_full_matrix.cc`, we use `boost::math::iround()` in a number of places. Unfortunately, this kills FreeBSD support:
```
In file included from /root/workspace/dealii/source/lac/lapack_full_ma…
-
Hi Sebastian,
I can see that you already implemented the cumulant approximation of the 4RMD. Is it possible to use it with the binary?
Best regards,
Quan
quanp updated
8 years ago
-
Previously, the following input file would have ran a combination of DFJK and DiskJK calculations. Now it seems to run DiskJK calculations twice.
```
molecule acetone {
0 1
H 0.000000 2.136732…
-
I just realized the obvious when thinking about dynamic analyses. We have to rename the commands how the local matrices are assembled. Here is a short memo
- Eigenvalue analysis
- Modal analysis
- Dam…
-
Using the latest numpy package from pypi.anaconda.org/carlkl/simple it's possible to trigger a segfault with:
```
python -c "import numpy as np; X = np.random.randn(200, 200); print(np.linalg.svd(X))…
-
If want to compute TICA at two (or more) different taus, it should be possible to skip the first pass of
TICA once the mean is already known.
gph82 updated
8 years ago
-
I specified a model with random slopes for the three-way interaction of three factors but no slopes for the other main effects and interactions. However, when I fit this model I get a lot more random…
-
Say I want for some reason to calculate the reconstruction error of TICA.
The TICA class does not have a inverse_transform method implemented.
Looking at `_transform_array`
``` python
def _transform…