[wip] hea integration (for GAM fit)

.github/workflows/python-package.yml

@@ -23,17 +23,21 @@ jobs:
               uses: actions/setup-python@v3
               with:
                   python-version: ${{ matrix.python-version }}
+            - name: Install uv
+              uses: astral-sh/setup-uv@v8.2.0
             - name: Install dependencies
+              # uv (not pip) so the [tool.uv.sources] git override for hea is
+              # honored, at the exact commit pinned in uv.lock.
               run: |
-                  python -m pip install --upgrade pip
-                  python -m pip install flake8 pytest
-                  pip install -e .
+                  uv sync --locked --python ${{ matrix.python-version }}
+                  uv pip install flake8 pytest
             - name: Lint with flake8
               run: |
                   # stop the build if there are Python syntax errors or undefined names
-                  flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
+                  # (--extend-exclude=.venv: uv sync puts the venv in the workspace, keep it out of lint)
+                  uv run --no-sync flake8 . --extend-exclude=.venv --count --select=E9,F63,F7,F82 --show-source --statistics
                   # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
-                  flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
+                  uv run --no-sync flake8 . --extend-exclude=.venv --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
             - name: Test with pytest
               run: |
-                  pytest
+                  uv run --no-sync pytest

examples/B2-Zar-2010.ipynb

@@ -133,7 +133,7 @@
     }
    ],
    "source": [
-    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c1 = Circular(data=d1)\n",
     "\n",
     "# Figure 26.2\n",
@@ -176,7 +176,7 @@
    ],
    "source": [
     "d2 = load_data(\"D2\", source=\"zar\")\n",
-    "c2 = Circular(data=d2[\"θ\"].values[:], w=d2[\"w\"].values[:])\n",
+    "c2 = Circular(data=d2[\"θ\"].to_numpy()[:], w=d2[\"w\"].to_numpy()[:])\n",
     "\n",
     "# figure 26.4\n",
     "c2.plot(\n",
@@ -238,7 +238,7 @@
     "\n",
     "from pycircstat2.utils import data2rad\n",
     "\n",
-    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "n = len(d1)\n",
     "alpha = data2rad(d1, k=360)\n",
     "\n",
@@ -335,8 +335,8 @@
    ],
    "source": [
     "d2 = load_data(\"D2\", source=\"zar\")\n",
-    "alpha = np.asarray(data2rad(d2[\"θ\"].values[:], k=360))\n",
-    "w = d2[\"w\"].values[:]\n",
+    "alpha = np.asarray(data2rad(d2[\"θ\"].to_numpy()[:], k=360))\n",
+    "w = d2[\"w\"].to_numpy()[:]\n",
     "select = w != 0\n",
     "alpha = alpha[select]\n",
     "w = w[select]\n",
@@ -350,7 +350,7 @@
     "\n",
     "frame = pl.DataFrame(\n",
     "    {\n",
-    "        \"α\": d2[\"θ\"].values[select],\n",
+    "        \"α\": d2[\"θ\"].to_numpy()[select],\n",
     "        \"f\": w,\n",
     "        \"sin(α)\": sina,\n",
     "        \"f * sin(α)\": fsina,\n",
@@ -404,7 +404,7 @@
    "source": [
     "from scipy.stats import chi2\n",
     "\n",
-    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "n = len(d1)\n",
     "alpha = np.asarray(data2rad(d1, k=360))\n",
     "\n",
@@ -491,7 +491,7 @@
     }
    ],
    "source": [
-    "d3 = load_data(\"D3\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d3 = load_data(\"D3\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "alpha = np.asarray(data2rad(d3, k=360))\n",
     "alpha2 = 2 * alpha % (2 * np.pi)\n",
     "sin2a = np.sin(alpha2)\n",
@@ -569,13 +569,13 @@
       "│ ---    ┆ --- ┆ ---    ┆ ---       ┆ ---       │\n",
       "│ i64    ┆ i64 ┆ f64    ┆ f64       ┆ f64       │\n",
       "╞════════╪═════╪════════╪═══════════╪═══════════╡\n",
-      "│ 1      ┆ 160 ┆ 0.8954 ┆ -0.841401 ┆ 0.306245  │\n",
-      "│ 2      ┆ 169 ┆ 0.7747 ┆ -0.760467 ┆ 0.14782   │\n",
-      "│ 3      ┆ 117 ┆ 0.4696 ┆ -0.213194 ┆ 0.418417  │\n",
-      "│ 4      ┆ 140 ┆ 0.8794 ┆ -0.673659 ┆ 0.565267  │\n",
-      "│ 5      ┆ 186 ┆ 0.3922 ┆ -0.390051 ┆ -0.040996 │\n",
-      "│ 6      ┆ 134 ┆ 0.6952 ┆ -0.482926 ┆ 0.500085  │\n",
-      "│ 7      ┆ 171 ┆ 0.3338 ┆ -0.32969  ┆ 0.052218  │\n",
+      "│ 0      ┆ 160 ┆ 0.8954 ┆ -0.841401 ┆ 0.306245  │\n",
+      "│ 1      ┆ 169 ┆ 0.7747 ┆ -0.760467 ┆ 0.14782   │\n",
+      "│ 2      ┆ 117 ┆ 0.4696 ┆ -0.213194 ┆ 0.418417  │\n",
+      "│ 3      ┆ 140 ┆ 0.8794 ┆ -0.673659 ┆ 0.565267  │\n",
+      "│ 4      ┆ 186 ┆ 0.3922 ┆ -0.390051 ┆ -0.040996 │\n",
+      "│ 5      ┆ 134 ┆ 0.6952 ┆ -0.482926 ┆ 0.500085  │\n",
+      "│ 6      ┆ 171 ┆ 0.3338 ┆ -0.32969  ┆ 0.052218  │\n",
       "└────────┴─────┴────────┴───────────┴───────────┘\n",
       "k = 7; n = 10\n",
       "∑X = -3.69139; ∑Y = 1.94906\n",
@@ -598,16 +598,16 @@
    ],
    "source": [
     "d4 = load_data(\"D4\", source=\"zar\")\n",
-    "ms = np.asarray(data2rad(d4[\"mean\"].values[:], k=360))\n",
-    "rs = d4[\"r\"].values[:]\n",
+    "ms = np.asarray(data2rad(d4[\"mean\"].to_numpy()[:], k=360))\n",
+    "rs = d4[\"r\"].to_numpy()[:]\n",
     "X = rs * np.cos(ms)\n",
     "Y = rs * np.sin(ms)\n",
-    "i = d4.index.values[:]\n",
+    "i = np.arange(len(d4))\n",
     "\n",
     "frame = pl.DataFrame(\n",
     "    {\n",
     "        \"Sample\": i,\n",
-    "        \"μ\": d4[\"mean\"].values,\n",
+    "        \"μ\": d4[\"mean\"].to_numpy(),\n",
     "        \"r\": rs,\n",
     "        \"X\": X,\n",
     "        \"Y\": Y,\n",
@@ -676,7 +676,7 @@
     }
    ],
    "source": [
-    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c1 = Circular(data=d1)\n",
     "\n",
     "n = c1.n\n",
@@ -735,7 +735,7 @@
     }
    ],
    "source": [
-    "d7 = load_data(\"D7\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d7 = load_data(\"D7\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c7 = Circular(data=d7)\n",
     "\n",
     "n = c7.n\n",
@@ -790,7 +790,7 @@
     }
    ],
    "source": [
-    "d7 = load_data(\"D7\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d7 = load_data(\"D7\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c7 = Circular(data=d7)\n",
     "\n",
     "if c7.mean_lb < np.deg2rad(90) < c7.mean_ub:\n",
@@ -837,7 +837,7 @@
     }
    ],
    "source": [
-    "d8 = load_data(\"D8\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d8 = load_data(\"D8\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c8 = Circular(data=d8)\n",
     "alpha = c8.alpha\n",
     "n = len(alpha)\n",
@@ -918,7 +918,7 @@
     {
      "data": {
       "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x11d56ee90>]"
+       "[<matplotlib.lines.Line2D at 0x119d45450>]"
       ]
      },
      "execution_count": 19,
@@ -937,7 +937,7 @@
     }
    ],
    "source": [
-    "d8 = load_data(\"D8\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d8 = load_data(\"D8\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c8 = Circular(data=d8)\n",
     "alpha = c8.alpha\n",
     "n = len(alpha)\n",
@@ -987,7 +987,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "d9 = load_data(\"D9\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d9 = load_data(\"D9\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c9 = Circular(data=d9)\n",
     "\n",
     "med = c9.median\n",
@@ -1033,8 +1033,8 @@
    ],
    "source": [
     "d10 = load_data(\"D10\", source=\"zar\")\n",
-    "s1 = d10[d10[\"sample\"] == 1][\"θ\"].values[:]\n",
-    "s2 = d10[d10[\"sample\"] == 2][\"θ\"].values[:]\n",
+    "s1 = d10.filter(pl.col(\"sample\") == 1)[\"θ\"].to_numpy()[:]\n",
+    "s2 = d10.filter(pl.col(\"sample\") == 2)[\"θ\"].to_numpy()[:]\n",
     "\n",
     "c10_s1 = Circular(s1)\n",
     "c10_s2 = Circular(s2)\n",
@@ -1079,9 +1079,9 @@
    ],
    "source": [
     "d11 = load_data(\"D11\", source=\"zar\")\n",
-    "s1 = d11[d11[\"sample\"] == 1][\"θ\"].values[:]\n",
-    "s2 = d11[d11[\"sample\"] == 2][\"θ\"].values[:]\n",
-    "s3 = d11[d11[\"sample\"] == 3][\"θ\"].values[:]\n",
+    "s1 = d11.filter(pl.col(\"sample\") == 1)[\"θ\"].to_numpy()[:]\n",
+    "s2 = d11.filter(pl.col(\"sample\") == 2)[\"θ\"].to_numpy()[:]\n",
+    "s3 = d11.filter(pl.col(\"sample\") == 3)[\"θ\"].to_numpy()[:]\n",
     "\n",
     "c11_s1 = Circular(s1)\n",
     "c11_s2 = Circular(s2)\n",
@@ -1130,8 +1130,8 @@
     "\n",
     "\n",
     "d12 = load_data(\"D12\", source=\"zar\")\n",
-    "s1 = d12[d12[\"sample\"] == 1][\"θ\"].values[:]\n",
-    "s2 = d12[d12[\"sample\"] == 2][\"θ\"].values[:]\n",
+    "s1 = d12.filter(pl.col(\"sample\") == 1)[\"θ\"].to_numpy()[:]\n",
+    "s2 = d12.filter(pl.col(\"sample\") == 2)[\"θ\"].to_numpy()[:]\n",
     "\n",
     "c12_s1 = Circular(s1)\n",
     "c12_s2 = Circular(s2)\n",
@@ -1163,10 +1163,10 @@
     "\n",
     "\n",
     "d13 = load_data(\"D13\", source=\"zar\")\n",
-    "s1 = d13[d13[\"sample\"] == 1][\"θ\"].values[:]\n",
-    "w1 = d13[d13[\"sample\"] == 1][\"w\"].values[:]\n",
-    "s2 = d13[d13[\"sample\"] == 2][\"θ\"].values[:]\n",
-    "w2 = d13[d13[\"sample\"] == 2][\"w\"].values[:]\n",
+    "s1 = d13.filter(pl.col(\"sample\") == 1)[\"θ\"].to_numpy()[:]\n",
+    "w1 = d13.filter(pl.col(\"sample\") == 1)[\"w\"].to_numpy()[:]\n",
+    "s2 = d13.filter(pl.col(\"sample\") == 2)[\"θ\"].to_numpy()[:]\n",
+    "w2 = d13.filter(pl.col(\"sample\") == 2)[\"w\"].to_numpy()[:]\n",
     "\n",
     "c13_s1 = Circular(data=s1, w=w1)\n",
     "c13_s2 = Circular(data=s2, w=w2)\n",
@@ -1208,8 +1208,8 @@
    ],
    "source": [
     "d12 = load_data(\"D12\", source=\"zar\")\n",
-    "c12_s1 = Circular(data=d12[d12[\"sample\"] == 1][\"θ\"].values[:])\n",
-    "c12_s2 = Circular(data=d12[d12[\"sample\"] == 2][\"θ\"].values[:])\n",
+    "c12_s1 = Circular(data=d12.filter(pl.col(\"sample\") == 1)[\"θ\"].to_numpy()[:])\n",
+    "c12_s2 = Circular(data=d12.filter(pl.col(\"sample\") == 2)[\"θ\"].to_numpy()[:])\n",
     "\n",
     "from pycircstat2.hypothesis import wheeler_watson_test\n",
     "\n",
@@ -1244,8 +1244,8 @@
    ],
    "source": [
     "d14 = load_data(\"D14\", source=\"zar\")\n",
-    "c14_s1 = Circular(data=d14[d14[\"sex\"] == \"male\"][\"θ\"].values[:])\n",
-    "c14_s2 = Circular(data=d14[d14[\"sex\"] == \"female\"][\"θ\"].values[:])\n",
+    "c14_s1 = Circular(data=d14.filter(pl.col(\"sex\") == \"male\")[\"θ\"].to_numpy()[:])\n",
+    "c14_s2 = Circular(data=d14.filter(pl.col(\"sex\") == \"female\")[\"θ\"].to_numpy()[:])\n",
     "\n",
     "from pycircstat2.hypothesis import wallraff_test\n",
     "\n",
@@ -1281,8 +1281,8 @@
    ],
    "source": [
     "d15 = load_data(\"D15\", source=\"zar\")\n",
-    "s1 = time2float(d15[d15[\"sex\"] == \"male\"][\"time\"].values[:])\n",
-    "s2 = time2float(d15[d15[\"sex\"] == \"female\"][\"time\"].values[:])\n",
+    "s1 = time2float(d15.filter(pl.col(\"sex\") == \"male\")[\"time\"].to_numpy()[:])\n",
+    "s2 = time2float(d15.filter(pl.col(\"sex\") == \"female\")[\"time\"].to_numpy()[:])\n",
     "\n",
     "c15_s1 = Circular(data=s1)\n",
     "c15_s2 = Circular(data=s2)\n",
@@ -1323,8 +1323,8 @@
     "\n",
     "d20 = load_data(\"D20\", source=\"zar\")\n",
     "\n",
-    "a = Circular(data=d20[\"Insect\"].values[:])\n",
-    "b = Circular(data=d20[\"Light\"].values[:])\n",
+    "a = Circular(data=d20[\"Insect\"].to_numpy()[:])\n",
+    "b = Circular(data=d20[\"Light\"].to_numpy()[:])\n",
     "\n",
     "from pycircstat2.correlation import circ_corrcc\n",
     "\n",
@@ -1358,8 +1358,8 @@
    "source": [
     "d21 = load_data(\"D21\", source=\"zar\")\n",
     "\n",
-    "a = Circular(data=d21[\"θ\"].values[:]).alpha\n",
-    "x = d21[\"X\"].values[:]\n",
+    "a = Circular(data=d21[\"θ\"].to_numpy()[:]).alpha\n",
+    "x = d21[\"X\"].to_numpy()[:]\n",
     "\n",
     "from pycircstat2.correlation import circ_corrcl\n",
     "\n",
@@ -1396,8 +1396,8 @@
    ],
    "source": [
     "d22 = load_data(\"D22\", source=\"zar\")\n",
-    "a = Circular(data=d22[\"evening\"].values[:])\n",
-    "b = Circular(data=d22[\"morning\"].values[:])\n",
+    "a = Circular(data=d22[\"evening\"].to_numpy()[:])\n",
+    "b = Circular(data=d22[\"morning\"].to_numpy()[:])\n",
     "res = circ_corrcc(a, b, test=True, method=\"nonparametric\")\n",
     "\n",
     "print(f\"r = {res.r:.5f}, reject? = {res.reject_null}\")"
@@ -1430,7 +1430,7 @@
    ],
    "source": [
     "d2 = load_data(\"D2\", source=\"zar\")\n",
-    "c2 = Circular(data=d2[\"θ\"].values[:], w=d2[\"w\"].values[:])\n",
+    "c2 = Circular(data=d2[\"θ\"].to_numpy()[:], w=d2[\"w\"].to_numpy()[:])\n",
     "\n",
     "from pycircstat2.hypothesis import chisquare_test\n",
     "\n",
@@ -1462,7 +1462,7 @@
     }
    ],
    "source": [
-    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].values[:]\n",
+    "d1 = load_data(\"D1\", source=\"zar\")[\"θ\"].to_numpy()[:]\n",
     "c1 = Circular(data=d1)\n",
     "\n",
     "from pycircstat2.hypothesis import watson_test\n",
@@ -1480,7 +1480,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Last updated: 2026-06-08 17:36:42 CEST\n",
+      "Last updated: 2026-06-09 13:17:14 CEST\n",
       "\n",
       "Python implementation: CPython\n",
       "Python version       : 3.13.11\n",