Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/tky823/bss-visualize
A visualization library for blind source separation
https://github.com/tky823/bss-visualize
Last synced: 24 days ago
JSON representation
A visualization library for blind source separation
- Host: GitHub
- URL: https://github.com/tky823/bss-visualize
- Owner: tky823
- Created: 2022-12-31T12:39:50.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-06-05T14:42:38.000Z (over 1 year ago)
- Last Synced: 2023-07-16T04:19:04.236Z (over 1 year ago)
- Language: Python
- Size: 260 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# bss-visualize
A visualization toolkit for blind source separation.
## Examples
### Plot
```python
import matplotlib.pyplot as plt
import numpy as np
from bss_visualize import plot_by_elapsed_time, plot_by_iteration, set_tex_style
params = {
"figure.dpi": 100,
"font.size": 20,
}
plt.rcParams.update(params)
set_tex_style()
data = {
"IP1": np.arange(1, 10),
"IP2": np.arange(3, 10),
"ISS1": np.arange(2, 10, 2),
"ISS2": np.arange(4, 12),
}
elapsed_time = {
"IP1": np.arange(9) / 5,
"IP2": np.arange(7),
"ISS1": np.arange(4),
"ISS2": np.arange(8) / 2,
}
labels = ["IP1", "IP2", "ISS1", "ISS2"]
# plot by iteration
fig, ax = plt.subplots(figsize=(12, 8))
handles = plot_by_iteration(
ax,
data,
labels=labels,
linestyle=["-", "-", "--", "--"],
marker=["o", "o", "s", "s"],
palette=["darkgray", "black", "darkgray", "black"],
)
ax.legend(handles, labels, bbox_to_anchor=(1.01, 1), loc="upper left", borderaxespad=0)
fig.savefig("SDRi-curve_by_iteration.pdf", bbox_inches="tight")
fig.savefig("SDRi-curve_by_iteration.png", bbox_inches="tight")
# plot by elapsed time
fig, ax = plt.subplots(figsize=(12, 8))
handles = plot_by_elapsed_time(ax, data, elapsed_time, labels=labels)
ax.legend(handles, labels, bbox_to_anchor=(1.01, 1), loc="upper left", borderaxespad=0)
fig.savefig("SDRi-curve_by_elapsed-time.pdf", bbox_inches="tight")
fig.savefig("SDRi-curve_by_elapsed-time.png", bbox_inches="tight")
```
### Group Boxplots
```python
import matplotlib.pyplot as plt
import numpy as np
from bss_visualize import set_tex_style, boxplot_by_group
params = {
"figure.dpi": 100,
"font.size": 20,
}
plt.rcParams.update(params)
set_tex_style()
sdri = {
"Gauss-ILRMA": {
"IP1": np.arange(0, 10),
"IP2": np.arange(5, 10),
"ISS1": np.arange(2, 12),
"ISS2": np.arange(-1, 5),
},
r"$t$-ILRMA"
+ "\n"
+ r"($\nu=10^{3}$)": {
"IP1": np.arange(1, 11),
"IP2": np.arange(6, 11),
"ISS1": np.arange(3, 13),
"ISS2": np.arange(0, 6),
},
"GGD-ILRMA"
+ "\n"
+ r"($\beta=1.95$)": {
"IP1": np.arange(-2, 8),
"IP2": np.arange(3, 8),
"ISS1": np.arange(0, 10),
"ISS2": np.arange(-3, 3),
},
}
labels = [
"Gauss-ILRMA",
r"$t$-ILRMA" + "\n" + r"($\nu=10^{3}$)",
"GGD-ILRMA" + "\n" + r"($\beta=1.95$)",
]
names = ["IP1", "IP2", "ISS1", "ISS2"]
fig, ax = plt.subplots(figsize=(12, 8))
handles = boxplot_by_group(ax, sdri, labels=labels, names=names)
ax.legend(handles, names, bbox_to_anchor=(1.01, 1), loc="upper left", borderaxespad=0)
ax.set_ylabel("SDR improvement [dB]")
fig.savefig("SDRi.pdf", bbox_inches="tight")
fig.savefig("SDRi.png", bbox_inches="tight")
```
