Deepcell-tracking

🐛 Bug Fixes

<details>
<summary>Speed up `trks_stats` by a factor of 5 willgraf (84)</summary>

`trks_stats` is used to calculate various statistics on a trks file. The old implementation ended up calling `np.unique` inside a double for loop, which caused the runtime to spike. These changes update the internal logic to greatly improve the runtime (namely, by separating the logic into two different for loops, updating the time complexity from O(batch * frames * cells) to O(batch*(frames+cells)).

Some example times on the current registry tracking data (all times include time to run `load_trks`):

- `test.trks` from 227s to 47s
- `val.trks` from 549s to 110s
- `train.trks` from 1968s to 386s
</details>

🐛 Bug Fixes

<details>
<summary>Use `np.any` instead of `np.sum` when finding all frames that contain a given object. willgraf (82)</summary>

`np.sum` counts all the frames the object is in, but we really only care that it exists in the given frame, not the number. `np.any` is about 3x faster (in my examples lineage), and should help speed up the slowest code in both `is_valid_lineage` and `relabel_sequential_lineage`.

Additionally, `is_valid_lineage` has some improved warning messages, and saves a single loop across the cells by iteratively removing the cells with the lineage from a set of all cells.
</details>


🧰 Maintenance

<details>
<summary>Bump the version to 0.5.1. willgraf (83)</summary>


</details>

🚀 Features

<details>
<summary>Improve error handling for `relabel_sequential_lineage` and `is_valid_lineage` willgraf (73)</summary>

`is_valid_lineage` only checks the lineage object itself. However, the lineage object is completely dependent on the `y` label mask. This function has been changed to require the `y` array, and is a much stronger validation of lineages. (Closes 72) This PR changes the `is_valid_lineage` signature, and should be considered a breaking change.

Additionally, `relabel_sequential` has been found to map unknown integers to 0. This caused some confusion that the original `.trk` file had a 0 as a daughter value. When this happens, `relabel_sequential_lineage` will warn the user that the value is getting mapped to 0. This will only happen with invalid lineages.

Finally, the `Track` object has been updated to only call `relabel_sequential_lineage` if the lineage is found to be valid. This should help reduce instantiation time by skipping the relabel step for invalid batches.

</details>

<details>
<summary>Migrate `Track` and `concat_tracks` to `deepcell.data.tracking`. willgraf (79)</summary>

These functions are being updated to use `tf.SparseTensors` to mitigate their memory footprint. However, `tensorflow` is not a dependency of `deepcell-tracking`, and is only used in the `deepcell` package. Fortunately, these functions are purely used in training pre-processing and are not actually used in this package.

This is a breaking change, to use the `Track` or `concat_tracks` functions, please use either `deepcell-tracking<0.5.0` or `deepcell>=0.11.0`.
</details>


🐛 Bug Fixes

<details>
<summary>Fix `norecursedirs` to ignore build directory willgraf (80)</summary>

Tests started failing due to a duplicate test event found in the `build` directory. This should be ignored in `pytest.ini`, but is an erroneous comma in the path. This PR fixes the `norecursedirs` setting.
</details>

<details>
<summary>Fix bad warning message in `is_valid_lineage` willgraf (77)</summary>

There is a bad label in one of the parent checks. This PR updates the message for accuracy.
</details>

<details>
<summary>Validate parent lineage field as well as daughters. willgraf (76)</summary>

73 updated `is_valid_lineage` to do better of all cells. However, there is no extra validation of the `parent` field, which is used in the `isbi_utils` functions.

This PR adds some basic validation of the `parent` lineage field to improve the validation of the lineage objects.
</details>




🧰 Maintenance

<details>
<summary>Drop support for Python 3.5 willgraf (74)</summary>

Python 3.5 has reached EOL. This PR removes support for it.
</details>

<details>
<summary>Bump version to `0.5.0`. willgraf (81)</summary>


</details>

🐛 Bug Fixes

<details>
<summary>Fix bad warning message in `is_valid_lineage` willgraf (77)</summary>

There is a bad label in one of the parent checks. This PR updates the message for accuracy.
</details>

<details>
<summary>Validate parent lineage field as well as daughters. willgraf (76)</summary>

73 updated `is_valid_lineage` to do better of all cells. However, there is no extra validation of the `parent` field, which is used in the `isbi_utils` functions.

This PR adds some basic validation of the `parent` lineage field to improve the validation of the lineage objects.
</details>

🐛 Bug Fixes

<details>
<summary>Validate parent lineage field as well as daughters. willgraf (76)</summary>

73 updated `is_valid_lineage` to do better of all cells. However, there is no extra validation of the `parent` field, which is used in the `isbi_utils` functions.

This PR adds some basic validation of the `parent` lineage field to improve the validation of the lineage objects.
</details>

🐛 Bug Fixes

<details>
<summary>`is_valid_lineage` should check if the daughter ID is valid. willgraf (70)</summary>

If a cell's daughter ID is not a valid track ID, then `is_valid_lineage` will crash due to a `KeyError`. Instead, it should catch this error and note the lineage as invalid.
</details>