Periodictable

----------------

New:

* Add energy dependence for rare earths (Lynn and Seeger, 1990).

Modified:

* Use complex b_c when computing the coherent cross section, leading to
correct values of sigma_c and sigma_i for materials with large absorption.
With this change the tabulated values for B[10] are now shown to be
self-consistent within a few percent.

Breaking changes:

* Neutron scattering factors are returned with one value for each wavelength
even for energy independent elements. Previous versions returned a scalar
if the returned value was identical for each wavelength.

See [README.rst](https://github.com/pkienzle/periodictable/blob/v1.6.0/README.rst) for complete history.

----------------

Known issues:

* Incoherent scattering computed for contrast matched mixture in D2O_sld(),
differs from the value that would be computed for a compound with the same
isotope proportions and density computed in neutron_sld(). This may change
in a future release.

Breaking changes:

* Fix calculation of contrast match points for biomolecules. The old
formula used the density of H2O for the D2O sld calculation.
* Modify biomolecule support to use H[1] rather than T for labile hydrogen.
This will result in less error when the labile formula is used in lieu
of the natural formula or the contrast-matched formula, and make it more
obvious from glancing at the formula that labile hydrogen is present.
* Modify *fasta.Molecule* attributes, dropping *Hmass* and *Hsld*. *Hnatural*
has been moved to *natural_formula*. The formula with labile hydrogen is
stored in *labile_formula*, as well as *formula* as before.

New:

* Add *replace()* method to formula to allow isotope substitution.
* Add *nsf.D2O_match()* and *nsf.D2O_sld()* functions.

Modified:

* Neutron wavelength now defaults to 1.798 A when wavelength and energy are
both None in *neutron_sld()* and *neutron_scattering()* rather than
throwing an assertion error.
* *table* can be passed to neutron sld calculators as the source of isotope
information when parsing the chemical formula.
* Switch unit test framework from nose to pytest.
* Update docs.

See [README.rst](https://github.com/pkienzle/periodictable/blob/v1.5.3/README.rst) for complete history.

----------------

Modified:

* Carbon density changed from 2.1 to 2.2 to match CXRO, CRC and RSC. The NIST
X-ray attenuation tables use 2.26; the Handbook of Mineralogy has 2.09-2.23.
The Neutron Data Booklet gave the value as 1.9-2.3, and 2.1 was chosen
from this range. The remaining density will continue to use values from the
Neutron Data Booklet, which cites CRC as the primary source.
* Updated references.

See [README.rst](https://github.com/pkienzle/periodictable/blob/v1.5.2/README.rst) for complete history.

----------------

Modified:

* fasta uses natural abundance of H for biomolecule when computing the
D2O contrast match rather than the biomolecule with pure H[1].
* remove half-life units from column header in activation table since
each row gives its own units.

See [README.rst](https://github.com/pkienzle/periodictable/tree/v1.5.1/CHANGES.rst) for complete history.

----------------

New:

* mixture by mass and volume, e.g., 5 g NaCl // 50 mL H2O1
* multilayer materials, e.g., 5 um Si // 3 nm Cr // 8 nm Au
* add support for bio molecules with labile hydrogens
* update list of possible oxidation states to include rare states

Modified:

* fixed computation of incoherent cross section so it is consistent with
coherent cross section and total cross section

----------------

Modified:

* default density is now the isotopic density rather than the natural density