Peppy

Changed

- `pep_version` is no longer a required parameter to create a `peppy.Project` instance from a configuration file.

Fixed

- Performance issues during sample parsing. Two list comprehensions were combined to speed up this functionality.
- `KeyError` is thrown when attempting to access the `pep_version` of a `peppy.Project` instance instatiated from a sample table (`csv`)

Added

- Implementation of `__eq__` for the `peppy.Project` class such that two instances of the class can be compared using python's equality operators (`==`, `!=`).
- New `from_dict` function that lets a user instatiate a new `peppy.Project` object using an in-memory representation of a PEP (a `dict`). This supports database storage of PEPs.
- New `extended` flag for the `to_dict` method on `peppy.Project` objects. This creates a **richer** dictionary representation of PEPs.
- Better sample parsing

Changed

- Unify exceptions related to remote YAML file reading in `read_yaml` function. Now always a `RemoteYAMLError` is thrown.
- `Project` dict representation

Added

- Support for PEP `2.1.0`, whichi includes support for no YAML configuration file component (CSV only), automatic sample merging if there are any duplicates in sample table index column, and new project attributes: `sample_table_index` and `subsample_table_index`.

Fixed

- Project string representation; [Issue 368](https://github.com/pepkit/peppy/issues/368)

Fixed
- Bug with setuptools 58

Added

- Support for remote URL config files

Fixed

- Error when accessing `Project.subsample_table` property when no subsample tables were defined

Added

- `to_dict` method in `Sample` class that can include or exclude `Project` reference

Changed

- If there's just one `subsample_table` specified, `Project.subsample_table` property will return an object of `pandas.DataFrame` class rather than a `list` of ones

Fixed

- `TypeError` when `subsample_table` is set to `null`