readme.md file for generated_undersampled folder in sharepoint
Browse files- a3sm/readme.md +35 -0
a3sm/readme.md
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Generated Undersampled PCM Example Data
|
| 2 |
+
|
| 3 |
+
The files in this directory contain
|
| 4 |
+
some undersampled PCM data generated from selected example ground-truth data
|
| 5 |
+
in `photocurrent_data` folder.
|
| 6 |
+
These files are intended for quick testing and comparison of algorithms
|
| 7 |
+
so that every test can use the same undersampled data.
|
| 8 |
+
|
| 9 |
+
Currently the structure is as follows:
|
| 10 |
+
|
| 11 |
+
```
|
| 12 |
+
generated_undersampled
|
| 13 |
+
βββ <randomisation-scheme>_random
|
| 14 |
+
βββ <data-name>_<randomisation-scheme>
|
| 15 |
+
βββ trial_<trial-index>
|
| 16 |
+
βββ zero_filled
|
| 17 |
+
βββ <data-name>_sample_<sampling_percentage>_percent_coarse_J=<coarse-J>_<randomisation-scheme>_random.npy
|
| 18 |
+
```
|
| 19 |
+
|
| 20 |
+
Where:
|
| 21 |
+
|
| 22 |
+
- `<randomisation-scheme>` is one of `uniform` or `multilevel`
|
| 23 |
+
|
| 24 |
+
- `<data-name>` is the name of the ground-truth data,
|
| 25 |
+
for example `example_CIGS_256x256`
|
| 26 |
+
|
| 27 |
+
- `<trial-index>` is an integer index for different random trials
|
| 28 |
+
|
| 29 |
+
- `<sampling_percentage>` is the percentage of samples taken in the undersampling, for example `20.0` for 20%
|
| 30 |
+
|
| 31 |
+
- `<coarse-J>` is the number of coarse samples taken in the J dimension, for example `6`
|
| 32 |
+
|
| 33 |
+
|
| 34 |
+
The `.npy` files in the `zero_filled` folder contain the forward output of the generated undersampled PCM data.
|
| 35 |
+
The undersampled measurement PCM data can be computed from this using the inverse WHT.
|