int-brain-lab / ibleatools / 16464640454
45%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 16
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 48.495% (+10.8%) from 37.709%
16464640454

push

github

web-flow 
Added aggregation functions , reveal website (#22)

* Split compute_features into two functions

Different functions for computing features from pid vs files. Channel_labels are calculated now only in a single function(more cleaner).Added channel_labels to pandera schema for Ap-data. Added some utility functions for adding metadata to parquet files which can be used for bookkeeping later on.

* output_dir now defaults to None instead of current directory, and does not saves feature files

* Ruff and bug fix

* bugfix Allen not atlas _id

* Added handling of scratch dir. Now to be be stated as a parameter in compute_features

* Added the output directory structure in the documentation of the function setup_output_directory

* plot series

* plot histogram

* fix plot

* todo

* equal sign

* fallback std / 16 in case IQR is 0

* resort by channel

* Dartsort Parameters fix

* Ruff formatting

* Added aggregation functions

* uncomment the unit tests for feature computations

* add new features and start checking for regressions (#20)

* add new features and start checking for regressions

* change test fixture path - add data changelog

* hopefully this will fix the CI

* Added some basic unittests for aggregation functions. Fixed some minor bugs with the metadata updation part.

* reveal website figure 04: raw LFP

* reveal website: figure 03 histology

* load denoised

* Added aggregation methods for the features (#21)

* Added aggregation functions and their tests

* Changes made -
1. Made the concept of snippet_level_dir as a relative path, in reference to the output directory. Earlier it was absolute paths. And would have made it difficult, if the output datasets need to be moved to a different location or a computer.

2. Changed the extension of all the parquet files that were getting produced to .pqt  instead of .parquet to make it consistent with the previous repositiry.

3. Added different aggregation strategies for different features. spike_c... (continued)

725 of 1495 relevant lines covered (48.49%)

0.97 hits per line

Uncovered Existing Lines

Lines Coverage ∆ File
2
96.3
 -3.7% utils.py
10
96.03
 26.77% features.py
33
38.1
 38.1% data.py
55
0.0
 0.0% outliers.py
107
27.86
 27.86% plots.py
143
0.0
 0.0% reveal.py
175
30.33
 -11.54% feature_computation.py
Jobs
ID Job ID Ran Files Coverage
1 16464640454.1 32
48.49
 GitHub Action Run
Source Files on build 16464640454