plots for synthetic dataset

chr-peters · Jun 7, 2021 · 2c09316 · 2c09316
1 parent f5796bd
commit 2c09316
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 4 deletions.
diff --git a/notebooks/ratio_plots.ipynb b/notebooks/ratio_plots.ipynb
@@ -7,7 +7,7 @@
  "outputs": [],
  "source": [
  "from sketching import settings\n",
- "from sketching.datasets import Dataset, Covertype_Sklearn, KDDCup_Sklearn, Webspam_libsvm\n",
+ "from sketching.datasets import Dataset, Covertype_Sklearn, KDDCup_Sklearn, Webspam_libsvm, Synthetic_Dataset\n",
  "\n",
  "import pandas as pd\n",
  "\n",
@@ -80,7 +80,8 @@
  " titles = {\n",
  " \"covertype_sklearn\": \"Covertype\",\n",
  " \"kddcup_sklearn\": \"Kddcup\",\n",
- " \"webspam_libsvm_desparsed\": \"Webspam\"\n",
+ " \"webspam_libsvm_desparsed\": \"Webspam\",\n",
+ " \"synthetic_n_100000\": \"Synthetic\"\n",
  " }\n",
  "\n",
  " # add SGD\n",
@@ -140,6 +141,16 @@
  "dataset = Webspam_libsvm()\n",
  "make_plot(dataset, x_min=0, x_max=15000, y_min=1, y_max=2)"
  ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "dataset = Synthetic_Dataset(n_rows=100000)\n",
+ "make_plot(dataset, x_min=0, x_max=3000, y_min=1, y_max=35)"
+ ]
  }
  ],
  "metadata": {

diff --git a/notebooks/runtime_plots.ipynb b/notebooks/runtime_plots.ipynb
@@ -7,7 +7,7 @@
  "outputs": [],
  "source": [
  "from sketching import settings\n",
- "from sketching.datasets import Dataset, Covertype_Sklearn, KDDCup_Sklearn, Webspam_libsvm\n",
+ "from sketching.datasets import Dataset, Covertype_Sklearn, KDDCup_Sklearn, Webspam_libsvm, Synthetic_Dataset\n",
  "\n",
  "import pandas as pd\n",
  "\n",
@@ -77,7 +77,8 @@
  " titles = {\n",
  " \"covertype_sklearn\": \"Covertype\",\n",
  " \"kddcup_sklearn\": \"Kddcup\",\n",
- " \"webspam_libsvm_desparsed\": \"Webspam\"\n",
+ " \"webspam_libsvm_desparsed\": \"Webspam\",\n",
+ " \"synthetic_n_100000\": \"Synthetic\"\n",
  " }\n",
  "\n",
  " for cur_method in [\"l2s\", \"uniform\", \"sketching\"]:\n",
@@ -140,6 +141,17 @@
  "make_plot(dataset, x_min=None, x_max=None, y_min=1, y_max=2, sampling_time=False)\n",
  "make_plot(dataset, x_min=None, x_max=None, y_min=1, y_max=2, sampling_time=True)"
  ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "dataset = Synthetic_Dataset(n_rows=100000)\n",
+ "make_plot(dataset, x_min=None, x_max=None, y_min=None, y_max=None, sampling_time=False)\n",
+ "make_plot(dataset, x_min=None, x_max=None, y_min=None, y_max=None, sampling_time=True)"
+ ]
  }
  ],
  "metadata": {