Hi,
I am trying to load a CSV with no header using

df = vaex.open('data/star0000-1.csv',sep=",", header=None, error_bad_lines=False)

but I get

could not convert column 0, error: TypeError('getattr(): attribute name must be string'), will try to convert it to string
Giving up column 0, error: TypeError('getattr(): attribute name must be string')
could not convert column 

In Python 3 \> and \( are considered invalid escape sequences:

source:45: DeprecationWarning: invalid escape sequence \>
source:46: DeprecationWarning: invalid escape sequence \(

-- Docs: https://docs.pytest.org/en/latest/warnings.html

The solution is to remove the escape strings and declare the pattern as raw-string.

https://docs.scipy.org/doc/numpy/reference/ufuncs.html#ufunc

It would allow us to skip processing of masked-out elements without first compacting the arrays. It might only be applicable if the MaskedArray.content is Numpy, though.

In the docs, instead of images showing from the url, only the url is displayed. See here.

Will submit a PR . . .

When I run cooler cload -c1 1 -p1 2 -c2 4 - p2 5 .. on a file with four columns:

galactic        1601634 galactic        1125339
galactic        1601743 galactic        1135340
galactic        1597121 galactic        1126359
galactic        1598993 galactic        1135185
galactic        1597588 galactic        1125826

The error I get is:

Traceback (most recent call l

A user recently requested directions to functionality implemented by make_bounds() and was unaware of its existence because it is undocumented. This reminded me that many ncap2 functions listed by ncap2 -f (which few people know exists) may be undocumented. This task is to create a list of the undocumented functions, paste it here, and then to add a description of each to the User Guide befo

IDL> a=hdf_open(/create)

a=hdf_open(/create)
                   ^
% HDF_OPEN: Incorrect number of arguments.

GDL> a=hdf_open(/create)
% HDF_OPEN: Variable is undefined: <Undefined>
% Execution halted at: $MAIN$  

2) Feature Request

Currently the ordering of dimensions described in the schema is in many cases not listed in the documentation. E.g., for ElectricalSeries.data we should add to the docval that the dimensions are num_time | num_channels. This would help users avoid errors with ordering of dimensions.

This issue was motivated by #960

This issue is in part also related to #626

Hi, I would like to add support for OpenCV matrices (cv::Mat). All supported matrix classes in H5M* are templated based on scalar type (plus other things); cv::Mat, on the other hand, stores its datatype internally. I usually use a simple switch/case function to map OpenCV type constants to HDF5 constants and vice verse

You mention in the readme "We don't support the endian argument to createVariable. The h5py API does not appear to offer this feature."

h5py dose support endianness buts its a component of the dtype i.e. dtype='<f8' is little, while dtype='>f8' is big, you can use the newbyteorder function on a numpy dype object to set it like so

>>> np.dtype("<f, >f")
dtype([('f0', '<f4'), ('f1', '>f4

Its unclear under what the cost of compression is. It would be nice to be able to change the default compression threshold (or set it to infinity to turn it off). This would allow us to empirically determine when we want to do it and not to do it.

When the units are provided as attribute of an axis in a NXdata, I guess it would be quite easy to display them in addition to the label of the axis.

My use-case is quite simple: in SAXS, half of the software expect the scattering vector to be in 1/nm, while the other half expects them to be in 1/Å. Those units differ by a factor 10 so that the user is easily confused between the two.

The in

We currently ship our own set of filesystem functions to create / handle dirs and files on the OS we support.

Much of it could be simplified with C++ filesystem support, which is still not production ready in tooling and broad support.

On top of that, for MPI-parallel operations such as an mkdir -p for a series, we do implement safe but potentially meta-data-server "heavy" operations. We c

I was looking over the docs. From it a feature like writing NPZ-files was not directly obvious. It would be great to revise a bit the docs.

Hi! Hope this finds you well!

Is the LoomR tutorial not up-to-date anymore? Had a bug right in the tutorial: https://satijalab.org/loomR/loomR_tutorial.html

> library(loomR)
Loading required package: R6
Loading required package: hdf5r
> library(Seurat)
> hca <- connect(filename = "cc95ff89-2e68-4a08-a234-480eca21ce79.homo_sapiens.loom", mode = "r+")
> NormalizeData(object = hca, ch