The restrict and expand function hadf no overload for the Sample collection container.
This is an error in the interface that was made when I have copy-pasted the function.

The manual is now the only place where the ointerface is documented.

It is now possible to sonstruct the sample array from all thre containers.

Instead the std::vector is used.

I have added the part that allows to access a single dimension as a NumPy vector, reading and writing.
Now only the rehaul of the persentation and the conditions are left.

Now they also contain the new material about the functions for manipulating components.
Still missing are the new NumPy functions, the conditions and the rehaul of the presentation demo.

I have forgotten to add the reading fucntion.

It now supports a fucntion to apply the conditions to sampels and containers.
This means that the condition can set the value that are restricted.
It is equivalent in first restrict the samples and then expand them.
But no temporary needs to be created.

It is now also possible to optain an array that contains the value of the conditions not only the dimensions.

Updated PyYggdrasil to also support the fourth extension of the container interface for the SampleCollection.

Due to the simnplicity of some functions I have Implememented more fucntions directly in the header.

This function allows to set all elements of the array to a single value.

I have added the new function to the SampleList, such that the interaction is with NumPy is better.
This onbly requiered my to add one new fucntion.
i have allso updated the description.
I have also added the full fourth extension of the container interface for the SampleList to pyYggdrasil.
I have also fixed some typos in the description for the SampleArray.

The function to set the dimension from an Eigen vector was modified.
Now there are two overloads.
One of them takes a regular Eigen::VectorXd type, in an Eigen::Ref wrapper, as argument.
It is intended to be used from within C++.
A second overload takes the new vector definition I have just introduced.
This overload is intendet to be used from Python, since it follows the restrictions that are made by the pybind11 documentation.

It seams that Eigen is very logical.
When one defines a vector, the storage order must be such that they aligne with each other.
Meaning in a row major setting, the number of rows must be one and the number of columns dynamic.
In a column major setting they are swapped.
When I made the template I did not thought about that.

I have made the description of the types more clear, why we do it (because pybind11 says so).
I also have added a row vector to it.
This is also done for better integration with NumPy.

Not yet in the Python wrapper.
I have compiled it.

I have added also the histogram plot to the tree sampler plots.
I have deactivated it.

I have implemenbted the next extension in teh container interface for the array.
It compiles.