Tue Jan 9 10:22:53 CET 2024

dgerosa · Jan 9, 2024 · 1b7503b · 1b7503b
1 parent 1ef6027
commit 1b7503b
Showing 1 changed file with 36 additions and 13 deletions.
diff --git a/lectures/L07_numba_multiprocessing.ipynb b/lectures/L07_numba_multiprocessing.ipynb
@@ -45,7 +45,7 @@
     "- Write some core functions in compiled language like C or Fortran and interface it with python. This is also useful to recycle some legacy code written by the supervisor of your supervisor who doesn't know python.\n",
     "- Use a library that converts python into compiled code (with some restrictions)\n",
     "\n",
-    "Some libraries:"
+    "Some examples:"
    ]
   },
   {
@@ -82,7 +82,7 @@
     "    compiled C code. The advantage here is that the code looks like python, with some\n",
     "    declarations of the variable types with `cdef`.  Performance can be\n",
     "    really great when you need to explicitly write out loops over\n",
-    "    numpy array indices.\n",
+    "    numpy array indices. Advantage: you can identify the bottleneck of your code and cytonize only that part. \n",
     "  * [Numba](https://numba.pydata.org/) : this is a just-in-time\n",
     "    compiler.  It just requires a simple decorator and then it will\n",
     "    compile a python function the first time it is encountered.\n",
@@ -110,12 +110,38 @@
     "\n",
     "A python decorator is a function that takes another function as input, modifies it, and returns the modified function to the user. I realize that this sentence sounds tricky, but it's not. \n",
     "\n",
+    "```\n",
+    "def decorator(function):\n",
+    "    [do something]\n",
+    "    output=function(input)\n",
+    "    [do something else]\n",
+    "    return [result]\n",
+    "```\n",
+    "\n",
+    "Then if you have any function\n",
+    "\n",
+    "```\n",
+    "def myprecious(ring):\n",
+    "    return gollum\n",
+    "```       \n",
+    " \n",
+    "You can redefine it to e.g.\n",
+    "\n",
+    "myprecious=decorator\n",
+    "\n",
+    "\n",
+    "    \n",
+    "    \n",
+    "\n",
+    "\n",
+    "\n",
+    "\n",
     "Remember that in python everything is an object. Functions are objects, and classes are objects too. For instance, let's take this simple function:"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 48,
+   "execution_count": 1,
    "id": "e38aba41",
    "metadata": {},
    "outputs": [],
@@ -134,7 +160,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 28,
+   "execution_count": 2,
    "id": "efe340f1",
    "metadata": {},
    "outputs": [],
@@ -163,7 +189,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 29,
+   "execution_count": 3,
    "id": "0806aae1",
    "metadata": {
     "colab": {
@@ -174,14 +200,11 @@
    },
    "outputs": [
     {
-     "ename": "NameError",
-     "evalue": "name 'hello' is not defined",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
-      "Cell \u001b[0;32mIn[29], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m whello \u001b[38;5;241m=\u001b[39m make_sure(\u001b[43mhello\u001b[49m)\n\u001b[1;32m      2\u001b[0m whello()\n",
-      "\u001b[0;31mNameError\u001b[0m: name 'hello' is not defined"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "are you sure you want to greet the world? [y/n]y\n",
+      "Hello world\n"
      ]
     }
    ],