|
@@ -10,6 +10,10 @@ There are very few constraints on the dataset, which are discussed below. [Our p
|
|
|
|
|
|
## Installation
|
|
|
|
|
|
+You should have MATLAB installed, in particular we recommend
|
|
|
+**MATLAB R2023a.** as the tool was tested against this version.
|
|
|
+Octave will **not** work.
|
|
|
+
|
|
|
First, download or clone this repo.
|
|
|
|
|
|
Secondly, **install the [kmeans_opt](https://it.mathworks.com/matlabcentral/fileexchange/65823-kmeans_opt) package**, by either moving its main module or adding it to MATLAB's search path.
|
|
@@ -29,7 +33,7 @@ This is a breakdown of what each step entails:
|
|
|
|
|
|
3. Lastly, re-perform clustering on the newly found BMs. Run `AssignModulesAndQuantify.m`.
|
|
|
|
|
|
-4. Optionally, extract the sequence length for each body module of interest. After being propmted a directory for the dataset of interest, it will create a directory
|
|
|
+4. Optionally, extract the sequence length for each body module of interest. After being prompted a directory for the dataset of interest, it will create a directory
|
|
|
containing all sequencing grouped by experimental condition.
|
|
|
|
|
|
### Dataset format
|
|
@@ -40,11 +44,11 @@ Please make sure your dataset follows the following format:
|
|
|
|
|
|
- Each subject folder follows a purely numeric code, e.g. "123" or "007"; note that "007" and "7" are distinct.
|
|
|
|
|
|
-- Each subject folder contains the same set of evaluating condition, named `${SUBJECT}_{CONDITION}_DLC_3D.csv`, for example [058_OF_DLC_3D.csv](https://github.com/gchelini87/SEB3R/blob/main/WN%20results/058/058_OF_DLC_3D.csv). These will be processed in lexicographical order, therefore if you wish to ensure consistent ordering we recommend prefixing them with a number, e.g. `1_Control`, `2_ExperimentalCondition` etc.
|
|
|
+- Each subject folder contains the same set of evaluating condition, named `${SUBJECT}_{CONDITION}_DLC_3D.csv`, for example [058_OF_DLC_3D.csv](https://gin.g-node.org/gchelini/SEB3R_pipeline/src/master/WN%20results/058/058_OF_DLC_3D.csv). These will be processed in lexicographical order, therefore if you wish to ensure consistent ordering we recommend prefixing them with a number, e.g. `1_Control`, `2_ExperimentalCondition` etc.
|
|
|
|
|
|
- Each condition is stored in a CSV matrix as returned by DLC. In particular:
|
|
|
|
|
|
- - Each matrix should have the same header as shown [here](https://github.com/gchelini87/SEB3R/blob/main/WN%20results/058/058_OF_DLC_3D.csv)
|
|
|
+ - Each matrix should have the same header as shown [here](https://gin.g-node.org/gchelini/SEB3R_pipeline/src/master/WN%20results/058/058_OF_DLC_3D.csv)
|
|
|
|
|
|
- Each matrix should have the same number of frames and tracked body parts. **Warning: currently the number of followed body parts is 6 due to instrumental and visual occlusion limitations**. This may change in future releases!
|
|
|
|