Refactor signal visualizations to use Matplotlib, removing Altair dependencies and updating figure configurations for clarity

Merge branch 'main' into cours
Update workflow configuration and replace Altair with Matplotlib for signal visualizations
2025-02-27 18:33:19 +01:00 · 2025-02-27 17:52:01 +01:00 · 2025-02-27 17:32:53 +01:00 · 2025-02-10 18:36:13 +01:00 · 2025-02-10 18:36:10 +01:00
6 changed files with 173 additions and 134 deletions
--- a/.forgejo/workflows/serve.yaml
+++ b/.forgejo/workflows/serve.yaml
@ -9,12 +9,12 @@ jobs:
      - name: Checkout repository
        uses: actions/checkout@v4
      - name: Initialize virtual environment
-        run: /usr/bin/python -m venv .env
+        run: /usr/bin/python -m venv .venv
      - name: Install dependencies
-        run: ./.env/bin/pip install mystmd jupyter jupyterlab_myst ipykernel altair pandas
+        run: ./.venv/bin/pip install -r requirements.txt
      - name: Build static HTML
        run: |
-          . .env/bin/activate
+          . .venv/bin/activate
          myst build --execute --html
      - name: Copy files
        run: |
--- a/cours/SIN/02-signaux.md
+++ b/cours/SIN/02-signaux.md
@ -32,32 +32,26 @@ et un niveau **bas** ("Low").
 :label: fig:sig-logique
 ```{code-cell} python
 :tags: [remove-input]
-import altair as alt
+import matplotlib.pyplot as plt
+from matplotlib import ticker
 import numpy as np
-import pandas as pd

 rng = np.random.default_rng(25)

 n = 16
 t = np.arange(n+1)
-s = rng.choice([0, 1], n+1)
-s[-1] = s[-2]
-data = pd.DataFrame({
-  "t": t,
-  "s": s,
-})
-alt.Chart(
-  data
-).mark_line(
-  interpolate="step-after",
-  strokeWidth=3,
-).encode(
-  alt.X("t:Q").axis(title="Temps (s)").scale(domain=(0,n)),
-  alt.Y("s:Q", axis=alt.Axis(title="Signal logique", values=[0, 1], format=".0f")).scale(domain=(0,1)),
-).properties(
-  width="container",
-  height=100,
+s = rng.choice([0, 1], n)
+
+fig, ax = plt.subplots()
+ax.stairs(s, t, lw=3, baseline=None)
+ax.set(
+  xlim=(0, n),
+  ylim=(-.5, 1.5),
+  xlabel="Temps (s)",
+  ylabel="Signal logique",
 )
+ax.yaxis.set_major_locator(ticker.MultipleLocator(1))
+ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
 ```
 Exemple de signal logique
 ````
@ -77,35 +71,37 @@ Un exemple de signal analogique est donné en @fig:sig-analogique.
 :label: fig:sig-analogique
 ```{code-cell} python
 :tags: [remove-input]
-import altair as alt
+import matplotlib.pyplot as plt
+from matplotlib import ticker
 import numpy as np
-import pandas as pd
+from scipy.interpolate import CubicSpline
+from scipy.stats import qmc
+

 rng = np.random.default_rng(25)

 n = 20
 t_max = 16

-t = np.linspace(0, t_max, n)
-
+t_base = np.linspace(0, t_max, n)
+lhs = (qmc.LatinHypercube(d=n-2, rng=rng).random(1)[0] - .5) * t_max/n
+t = t_base + np.concatenate(([0], lhs, [0]))
+t = t_base
 s = 5 * rng.random(n)
-s[-1] = s[-2]
-data = pd.DataFrame({
-  "t": t,
-  "s": s,
-})
-alt.Chart(
-  data
-).mark_line(
-  interpolate="basis",
-  strokeWidth=3,
-).encode(
-  alt.X("t:Q").axis(title="Temps (s)").scale(domain=(0,t_max)),
-  alt.Y("s:Q", axis=alt.Axis(title="Signal analogique")).scale(domain=(0,5)),
-).properties(
-  width="container",
-  height=200,
+
+t_interp = np.linspace(0, t_max, 1024)
+s_interp = np.clip(CubicSpline(t, s)(t_interp), 0, 5)
+
+fig, ax = plt.subplots()
+ax.plot(t_interp, s_interp, lw=3)
+ax.set(
+  xlim=(0, t_max),
+  ylim=(-.5, 5.5),
+  xlabel="Temps (s)",
+  ylabel="Signal analogique",
 )
+ax.yaxis.set_major_locator(ticker.MultipleLocator(1))
+ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
 ```
 Exemple de signal analogique
 ````
@ -119,32 +115,26 @@ Un exemple de signal analogique est donné en @fig:sig-numerique.
 :label: fig:sig-numerique
 ```{code-cell} python
 :tags: [remove-input]
-import altair as alt
+import matplotlib.pyplot as plt
+from matplotlib import ticker
 import numpy as np
-import pandas as pd

 rng = np.random.default_rng(25)

 n = 16
 t = np.arange(n+1)
-s = rng.integers(0, 16, n+1)
-s[-1] = s[-2]
-data = pd.DataFrame({
-  "t": t,
-  "s": s,
-})
-alt.Chart(
-  data
-).mark_line(
-  interpolate="step-after",
-  strokeWidth=3,
-).encode(
-  alt.X("t:Q").axis(title="Temps (s)").scale(domain=(0,n)),
-  alt.Y("s:Q", axis=alt.Axis(title="Signal numérique", values=np.arange(0, 16))).scale(domain=(0,15)),
-).properties(
-  width="container",
-  height=200,
+s = rng.integers(0, 16, n)
+
+fig, ax = plt.subplots()
+ax.stairs(s, t, lw=3, baseline=None)
+ax.set(
+  xlim=(0, n),
+  ylim=(-.5, 16.5),
+  xlabel="Temps (s)",
+  ylabel="Signal numérique",
 )
+ax.yaxis.set_major_locator(ticker.MultipleLocator(1))
+ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
 ```
 Exemple de signal numérique
 ````
--- a/cours/SIN/03-can.md
+++ b/cours/SIN/03-can.md
@ -37,69 +37,64 @@ La **caractéristique** du CAN est la courbe représentant la valeur numérique
 :label: fig:exemple-can
 ```{code-cell} python
 :tags: [remove-input]
-import altair as alt
+import matplotlib.pyplot as plt
+from matplotlib import ticker
 import numpy as np
-import pandas as pd
-from scipy import interpolate
+from scipy.interpolate import CubicSpline
+from scipy.stats import qmc

-rng = np.random.default_rng(25)
+rng = np.random.default_rng(50)

 n = 20
-t_max = 16
+t_max = 8
+n_interp = t_max * 100 + 1

-T = np.linspace(0, t_max, 1601)
-y = np.clip(
-    interpolate.BSpline(np.linspace(0, t_max, n), 5 * rng.random(n), 2)(T),
-    0,
-    5,
+t_base = np.linspace(0, t_max, n)
+lhs = (qmc.LatinHypercube(d=n-2, rng=rng).random(1)[0] - .5) * t_max/n
+t = t_base + np.concatenate(([0], lhs, [0]))
+t = t_base
+s = 5 * rng.random(n)
+
+t_interp = np.linspace(0, t_max, n_interp)
+s_interp = np.clip(CubicSpline(t, s)(t_interp), 0, 5)
+s_n = np.full_like(t_interp[::50], np.nan)
+s_n = np.floor(s_interp[::50] * 8 / 5)
+s_n[s_n == 8] = 7
+
+fig, ax = plt.subplots()
+ax2 = ax.twinx()
+
+ax.plot(t_interp, s_interp, lw=3)
+ax2.scatter(t_interp[::50], s_n, color="C1")
+
+ax.grid(False, axis="y")
+ax.grid(True, axis="x", which="both")
+ax2.grid(True)
+
+ax.set(
+  xlim=(0, t_max),
+  ylim=(-.5, 5.5),
+  xlabel="Temps (s)",
 )
-y_n = np.full([1601], np.nan)
-y_n[::50] = np.floor(y[::50] * 8 / 5)
-y_n[y_n == 8] = 7
-
-
-data = pd.DataFrame({
-  "t": T,
-  "s": y,
-  "s_n": y_n,
-})
-
-base = alt.Chart(
-  data
-).encode(
-  alt.X("t:Q").axis(title="Temps (s)").scale(domain=(0,t_max)),
+ax.set_ylabel("Signal analogique (V)", color="C0")
+ax2.set(
+  ylim=(-8/5*.5, 8/5*5.5),
 )
+ax2.set_ylabel("Signal numérique", color="C1")

-ch = base.mark_line(
-  interpolate="basis",
-  strokeWidth=3,
-  color="#6666cc",
-).encode(
-  alt.Y("s:Q", axis=alt.Axis(title="Signal analogique", titleColor="#6666cc")).scale(domain=(0,5)),
-)
+ax.yaxis.set_major_locator(ticker.MultipleLocator(1))
+ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
+ax.xaxis.set_minor_locator(ticker.MultipleLocator(.5))
+ax2.set_yticks(np.arange(9), np.concatenate((np.arange(8), [""])))

-ch_n = base.mark_point(
-  filled=True,
-  color="#ff6600",
-).encode(
-  alt.Y(
-    "s_n:Q",
-    axis=alt.Axis(
-      title="Signal numérisé",
-      titleColor="#ff6600",
-      values=np.arange(8),
-    )
-  ).scale(domain=(0,8)),
-)
-
-alt.layer(ch_n, ch).resolve_scale(
-  y="independent",
-).properties(
-  width="container",
-  height=200,
-)
+arr = ax2.annotate("", xy=(0, 0), xytext=(0.5, 0), arrowprops=dict(arrowstyle="<->"))
+ax2.annotate("$T_e$", (0.5, 1), xycoords=arr, ha="center", va="bottom")

+arr2 = ax2.annotate("", xy=(0.5, 0), xytext=(0.5, 1), arrowprops=dict(arrowstyle="<->"))
+ax2.annotate("$q$", (1, 0.5), xycoords=arr2, ha="left", va="center")

+arr3 = ax2.annotate("", xy=(1, 0), xytext=(1, 8), arrowprops=dict(arrowstyle="<->"))
+ax2.annotate("$V_{pe}$", (1, 0.5), xycoords=arr3, ha="left", va="center")
 ```
 Signal analogique et signal numérisé.
 ````
@ -108,34 +103,37 @@ Signal analogique et signal numérisé.
 :label: fig:carac-can
 ```{code-cell} python
 :tags: [remove-input]
-import altair as alt
+import matplotlib.pyplot as plt
+from matplotlib import ticker
 import numpy as np
-import pandas as pd
-from scipy import interpolate

 N = 8
-s_n = np.arange(N+1)
-s_n[-1] = s_n[-2]
-data = pd.DataFrame({
-  "s_n": s_n,
-  "s_a": np.linspace(0, 5, N+1),
-})
+s_n = np.arange(N)
+s_a = np.linspace(0, 5, N+1)

-alt.Chart(
-  data
-).mark_line(
-  interpolate="step-after",
-  strokeWidth=3,
-  color="#ff6600",
-).encode(
-  alt.X("s_a:Q").axis(title="Signal Analogique").scale(domain=(0,5)),
-  alt.Y("s_n:Q", axis=alt.Axis(title="Signal numérique", values=np.arange(N))).scale(domain=(0,N)),
-).properties(
-  width=200,
-  height=200,
+fig, ax = plt.subplots()
+
+ax.stairs(s_n, s_a, color="C1", lw=3, baseline=None)
+
+ax.set(
+  xlim=(0, 5),
+  ylim=(-1, N),
+  yticks=s_n,
+  xlabel="Signal analogique (V)",
+  ylabel="Signal numérique",
 )
+ax.set_xticks(s_a, [f"{v:.3f}" for v in s_a], rotation=45, ha="right", rotation_mode="anchor")
+ax.set_aspect(5/8, 'box')

+arr4 = ax.annotate(
+    "", xy=(s_a[0], 0), xytext=(s_a[1], 0), arrowprops=dict(arrowstyle="<->")
+)
+ax.annotate("$q$", (0.5, 1), xycoords=arr4, ha="center", va="bottom")

+arr5 = ax.annotate(
+    "", xy=(s_a[0], 1.5), xytext=(s_a[-1], 1.5), arrowprops=dict(arrowstyle="<->")
+)
+ax.annotate("$V_{pe}$", (0.5, 1), xycoords=arr5, ha="center", va="bottom")
 ```
 Caractéristique du CAN.
 ````
--- a/cours/SIN/matplotlibrc
+++ b/cours/SIN/matplotlibrc
@ -0,0 +1 @@
+../../matplotlibrc
--- a/43
+++ b/43
@ -0,0 +1,43 @@
+lines.linewidth: 3
+
+font.family: Fira Code
+
+image.cmap: inferno
+
+axes.linewidth: 1
+axes.grid: True
+axes.grid.which: major
+axes.titlelocation: right
+axes.titleweight: 700
+axes.axisbelow: True
+
+axes.prop_cycle: cycler(color=["#4269d0","#efb118","#ff725c","#6cc5b0","#3ca951","#ff8ab7","#a463f2","#97bbf5","#9c6b4e","#9498a0"])
+
+axes.formatter.use_locale: True
+
+grid.color: "#bebebe"
+grid.linewidth: 1
+grid.alpha: 1
+
+hatch.linewidth: 8
+hatch.color: "#00000013"
+
+boxplot.showmeans: true
+boxplot.meanprops.markeredgecolor: "k"
+boxplot.meanprops.marker: "+"
+boxplot.flierprops.markerfacecolor: C0
+boxplot.medianprops.color: C0
+
+figure.figsize: 8, 4.5
+figure.dpi: 96
+figure.constrained_layout.use: True
+
+xtick.direction: in
+xtick.major.size: 4
+xtick.minor.size: 2
+
+ytick.direction: in
+ytick.major.size: 4
+ytick.minor.size: 2
+
+savefig.format: pdf
--- a/requirements.txt
+++ b/requirements.txt
@ -0,0 +1,7 @@
+mystmd
+jupyter-server
+ipykernel
+matplotlib
+numpy
+pandas
+scipy
Author	SHA1	Message	Date
Edgar P. Burkhart	bc9d8904e2	Refactor signal visualizations to use Matplotlib, removing Altair dependencies and updating figure configurations for clarity	2025-02-27 18:33:19 +01:00
Edgar P. Burkhart	b032bb8610	Merge branch 'main' into cours	2025-02-27 17:52:01 +01:00
Edgar P. Burkhart	0503034e40	Update workflow configuration and replace Altair with Matplotlib for signal visualizations	2025-02-27 17:32:53 +01:00
Edgar P. Burkhart	6cc2200683	Merge branch 'cours'	2025-02-10 18:36:13 +01:00
Edgar P. Burkhart	51fe7460bf	Merge branch 'dev'	2025-02-10 18:36:10 +01:00