JRL | Joint Representation Learning for Top-N Recommendation | Machine Learning library

I am getting an error while running this query. Error : ORA-00937 not a single-group group function.

My code is

When I run the first Google API url with origin, destination and waypoints, it returns

https://maps.googleapis.com/maps/api/directions/json?origin=9.6614981,80.02554649999999&waypoints=6.9040322,79.948803&destination=6.053518500000001,6.053518500000001%7C&sensor=false

I'm trying to use the hugo-future-imperfect theme but I cannot make new Rmarkdown posts using: Addins > New Post > <>.

When I click done there is no file in the editor or in the filesystem, and the Console reports:

The documentation for the numpy.frombuffer function specifically says that the generated array will be one dimensional:

Interpret a buffer as a 1-dimensional array.

I'm not sure about the consequences of this quote. The documentation just tells me that the generated array will be one dimensional, but never says that the input buffer has to describe a one-dimensional object.

I have a (2D) Eigen matrix in C++. I would like to create a Python buffer which describes the content of the matrix. Then, I would like to use this buffer to somehow initialize my NumPy array and make it available to my python scripts. The goal is both to pass information to Python without copying data and to allow python modify the matrix (e.g. to initialize the matrix).

The C-API equivalent of numpy.frombuffer is PyArray_FromBuffer, and it also shares the single-dimension phrase, but it has more documentation (emphasis mine):

PyObject* PyArray_FromBuffer(PyObject* buf, PyArray_Descr* dtype, npy_intp count, npy_intp offset)

Construct a one-dimensional ndarray of a single type from an object, buf, that exports the (single-segment) buffer protocol (or has an attribute __buffer__ that returns an object that exports the buffer protocol). A writeable buffer will be tried first followed by a read- only buffer. The NPY_ARRAY_WRITEABLE flag of the returned array will reflect which one was successful. The data is assumed to start at offset bytes from the start of the memory location for the object. The type of the data in the buffer will be interpreted depending on the data- type descriptor, dtype. If count is negative then it will be determined from the size of the buffer and the requested itemsize, otherwise, count represents how many elements should be converted from the buffer.

Does "single-segment" mean that it cannot contain padding used, e.g., to align the rows of the matrix? In that case I'm screwed, because my matrix could very well use an alignment strategy that requires padding.

Back to the original question:

Is there a way for me to create a NumPy array which shares the memory with an pre-existing buffer?

Remark: there is a project on github called Eigen3ToPython, which aims at connecting eigen with python, but it does not allow for memory sharing (emphasis mine):

This library allows to: [...] Convert to/from Numpy arrays (np.array) in a transparent manner (however, memory is not shared between both representations)

EDIT Somebody might point out the similarly-titled question Numpy 2D- Array from Buffer?. Unfortunately, the solution given there does not seem to be a valid one for my case, because the generated 2D array does not share the memory with the original buffer.

EDIT: how is data organized in Eigen

Eigen maps 2D matrices in a 1D memory buffer by using strided access. A double precision 3x2 matrix, for instance, needs 6 double, i.e., 48 bytes. A 48-bytes buffer is allocated. The first element in this buffer represents the [0, 0] entry in the matrix.

In order to access the element [i, j], the following formula is used: