classification_report_pd

3springs · Oct 15, 2020 · 6cac624 · 6cac624
1 parent 195ffe3
commit 6cac624
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 44 deletions.
diff --git a/notebooks/c07_Recurrent_Neural_Networks/Recurrent_Neural_Networks.ipynb b/notebooks/c07_Recurrent_Neural_Networks/Recurrent_Neural_Networks.ipynb
@@ -4096,20 +4096,7 @@
     }
    },
    "outputs": [],
-   "source": [
-    "import pandas as pd\n",
-    "import sklearn.metrics\n",
-    "import numpy as np\n",
-    "\n",
-    "def classification_report(*args, **kwargs):\n",
-    "    out_df = pd.DataFrame(sklearn.metrics.classification_report(*args, **kwargs, output_dict=True)).T\n",
-    "    # Order cols\n",
-    "    out_df[[\"precision\",\"recall\",\"f1-score\",\"support\"]]  \n",
-    "    # Round\n",
-    "    out_df[[\"precision\",\"recall\",\"f1-score\"]]= out_df[[\"precision\",\"recall\",\"f1-score\"]].apply(lambda x: round(x,2))\n",
-    "    out_df[[\"support\"]]= out_df[[\"support\"]].apply(lambda x: x.astype(np.int))\n",
-    "    return out_df"
-   ]
+   "source": []
   },
   {
    "cell_type": "code",
@@ -4278,6 +4265,7 @@
     }
    ],
    "source": [
+    "from deep_ml_curriculum.classification_report import pd_classification_report\n",
     "df_report = classification_report(true, preds, labels=range(len(encoder.classes_)), target_names=encoder.classes_)\n",
     "df_report[df_report.support>0]"
    ]
@@ -5508,18 +5496,6 @@
     "1"
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2020-10-07T12:22:51.313045Z",
-     "start_time": "2020-10-07T12:22:51.218514Z"
-    }
-   },
-   "outputs": [],
-   "source": []
-  },
   {
    "cell_type": "code",
    "execution_count": 98,
@@ -5528,6 +5504,7 @@
      "end_time": "2020-10-07T12:58:49.315876Z",
      "start_time": "2020-10-07T12:55:35.018783Z"
     },
+    "lines_to_next_cell": 0,
     "scrolled": false
    },
    "outputs": [
@@ -6058,6 +6035,13 @@
    "outputs": [],
    "source": []
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -6121,7 +6105,12 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2020-10-07T12:22:51.313045Z",
+     "start_time": "2020-10-07T12:22:51.218514Z"
+    }
+   },
    "outputs": [],
    "source": []
   }
@@ -6152,7 +6141,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.9"
+   "version": "3.7.0"
   },
   "toc": {
    "base_numbering": 1,

diff --git a/notebooks/c07_Recurrent_Neural_Networks/Recurrent_Neural_Networks.py b/notebooks/c07_Recurrent_Neural_Networks/Recurrent_Neural_Networks.py
@@ -761,19 +761,6 @@ def training_loop(x_train, y_train, x_test, y_test, mode, epochs=1, bs=128, max_
 
 # This beats the baseline, so the model is doing better than the naive answer of "the same again". But lets break it down by lithology
 
-# +
-import pandas as pd
-import sklearn.metrics
-import numpy as np
-
-def classification_report(*args, **kwargs):
-    out_df = pd.DataFrame(sklearn.metrics.classification_report(*args, **kwargs, output_dict=True)).T
-    # Order cols
-    out_df[["precision","recall","f1-score","support"]]  
-    # Round
-    out_df[["precision","recall","f1-score"]]= out_df[["precision","recall","f1-score"]].apply(lambda x: round(x,2))
-    out_df[["support"]]= out_df[["support"]].apply(lambda x: x.astype(np.int))
-    return out_df
 
 
 # +
@@ -782,6 +769,7 @@ def classification_report(*args, **kwargs):
 # df_report[df_report.support>0]
 # -
 
+from deep_ml_curriculum.classification_report import pd_classification_report
 df_report = classification_report(true, preds, labels=range(len(encoder.classes_)), target_names=encoder.classes_)
 df_report[df_report.support>0]
 
@@ -977,9 +965,6 @@ def plot_well(df, model, depth_min=0, depth_max=18000, well_name="30_6-11", devi
 plot_well(df, model)
 confusion_matrix(true, preds)
 1
-# -
-
-
 # +
 # Params
 seq_length = 400  # CHANGE ME
@@ -1035,6 +1020,7 @@ def plot_well(df, model, depth_min=0, depth_max=18000, well_name="30_6-11", devi
 
 
 
+
 # ## Further Reading
 # - [Introduction to RNN](http://slazebni.cs.illinois.edu/spring17/lec02_rnn.pdf)
 # - [A friendly introduction to Recurrent Neural Networks](https://www.youtube.com/watch?v=UNmqTiOnRfg)