👌 Exact grid spacing, broaden mask and refactor tilling function

Final part of #52! Increasing number of training image tiles from 2111 to 2480! Couple of changes for that to work. The grid spacing -I flag in GMT now uses 250+e instead of 250 so that output cell size is precisely 250 metres. Mask_cell_radius was increased from 2 to 3 (i.e. 7x7 pixel neighbourhood mask) to get more tiles along the Siple Coast area. Tiling functions refactored for speed, fixing rounding errors, and linted using black. Note that we are still using the default gridline registration https://gmt.soest.hawaii.edu/doc/latest/GMT_Docs.html#gridline-registration when doing interpolation. Unit tests added for get_window_bounds and selective_tile functions!!
weiji14 · Nov 24, 2018 · ab0295e · ab0295e
1 parent 8706302
commit ab0295e
Show file tree

Hide file tree

Showing 2 changed files with 224 additions and 105 deletions.
diff --git a/data_prep.ipynb b/data_prep.ipynb
diff --git a/test_ipynb.ipynb b/test_ipynb.ipynb
@@ -147,6 +147,56 @@
       "    '0.05808/0.83244/0.02058/0.95071'\n",
       "ok\n",
       "Trying:\n",
+      "    xr.DataArray(\n",
+      "        data=np.zeros(shape=(36, 32)),\n",
+      "        coords={\"x\": np.arange(1, 37), \"y\": np.arange(1, 33)},\n",
+      "        dims=[\"x\", \"y\"],\n",
+      "    ).to_netcdf(path=\"/tmp/tmp_wb.nc\")\n",
+      "Expecting nothing\n",
+      "ok\n",
+      "Trying:\n",
+      "    get_window_bounds(filepath=\"/tmp/tmp_wb.nc\")\n",
+      "Expecting:\n",
+      "    Tiling: /tmp/tmp_wb.nc ... 2\n",
+      "    [(0.5, 4.5, 32.5, 36.5), (0.5, 0.5, 32.5, 32.5)]\n",
+      "ok\n",
+      "Trying:\n",
+      "    os.remove(\"/tmp/tmp_wb.nc\")\n",
+      "Expecting nothing\n",
+      "ok\n",
+      "Trying:\n",
+      "    xr.DataArray(\n",
+      "        data=np.random.RandomState(seed=42).rand(64).reshape(8, 8),\n",
+      "        coords={\"x\": np.arange(8), \"y\": np.arange(8)},\n",
+      "        dims=[\"x\", \"y\"],\n",
+      "    ).to_netcdf(path=\"/tmp/tmp_st.nc\", mode=\"w\")\n",
+      "Expecting nothing\n",
+      "ok\n",
+      "Trying:\n",
+      "    selective_tile(\n",
+      "       filepath=\"/tmp/tmp_st.nc\",\n",
+      "       window_bounds=[(1.0, 4.0, 3.0, 6.0), (2.0, 5.0, 4.0, 7.0)],\n",
+      "    )\n",
+      "Expecting:\n",
+      "    Tiling: /tmp/tmp_st.nc\n",
+      "    array([[[[0.18485446],\n",
+      "             [0.96958464]],\n",
+      "    <BLANKLINE>\n",
+      "            [[0.4951769 ],\n",
+      "             [0.03438852]]],\n",
+      "    <BLANKLINE>\n",
+      "    <BLANKLINE>\n",
+      "           [[[0.04522729],\n",
+      "             [0.32533032]],\n",
+      "    <BLANKLINE>\n",
+      "            [[0.96958464],\n",
+      "             [0.77513283]]]], dtype=float32)\n",
+      "ok\n",
+      "Trying:\n",
+      "    os.remove(\"/tmp/tmp_st.nc\")\n",
+      "Expecting nothing\n",
+      "ok\n",
+      "Trying:\n",
       "    xyz_data = 1000*pd.DataFrame(np.random.RandomState(seed=42).rand(60).reshape(20, 3))\n",
       "Expecting nothing\n",
       "ok\n",
@@ -166,26 +216,26 @@
       "Trying:\n",
       "    grid.to_array().values\n",
       "Expecting:\n",
-      "    array([[[388.9193 , 542.3901 , 711.7681 , 964.1099 , 798.63763],\n",
-      "            [361.2711 , 503.67337, 461.1001 , 207.78824, 272.23367],\n",
-      "            [358.95032, 346.20276, 130.3158 , 127.49137, 219.36589],\n",
-      "            [426.14886, 444.0474 , 185.34523, 382.60327, 479.42456],\n",
-      "            [495.67093, 731.8629 , 231.33981, 787.7948 , 808.39734]]],\n",
+      "    array([[[403.17618 , 544.92535 , 670.7824  , 980.75055 , 961.47723 ],\n",
+      "            [379.0757  , 459.26407 , 314.38297 , 377.78555 , 546.0469  ],\n",
+      "            [450.67664 , 343.26    ,  88.391594, 260.10492 , 452.3337  ],\n",
+      "            [586.09906 , 469.74008 , 216.8168  , 486.9802  , 642.2116  ],\n",
+      "            [451.4794  , 652.7244  , 325.77896 , 879.8973  , 916.7921  ]]],\n",
       "          dtype=float32)\n",
       "ok\n",
-      "4 items had no tests:\n",
+      "2 items had no tests:\n",
       "    data_prep\n",
-      "    data_prep.get_window_bounds\n",
       "    data_prep.parse_datalist\n",
-      "    data_prep.selective_tile\n",
-      "5 items passed all tests:\n",
+      "7 items passed all tests:\n",
       "   6 tests in data_prep.ascii_to_xyz\n",
       "   3 tests in data_prep.check_sha256\n",
       "   3 tests in data_prep.download_to_path\n",
       "   2 tests in data_prep.get_region\n",
+      "   3 tests in data_prep.get_window_bounds\n",
+      "   3 tests in data_prep.selective_tile\n",
       "   5 tests in data_prep.xyz_to_grid\n",
-      "19 tests in 9 items.\n",
-      "19 passed and 0 failed.\n",
+      "25 tests in 9 items.\n",
+      "25 passed and 0 failed.\n",
       "Test passed.\n"
      ]
     }