update rscript

fenke · Jul 19, 2024 · 640a9e6 · 640a9e6
1 parent f3a8fd3
commit 640a9e6
Show file tree

Hide file tree

Showing 6 changed files with 72 additions and 25 deletions.
diff --git a/corebridge/__init__.py b/corebridge/__init__.py
@@ -1 +1 @@
-__version__ = "0.2.15"
+__version__ = "0.2.16"
diff --git a/corebridge/rscript.py b/corebridge/rscript.py
@@ -5,11 +5,11 @@
            'check_script_inputs', 'check_script_output', 'generate_checksum_file', 'run_script']
 
 # %% ../nbs/02_rscriptbridge.ipynb 4
-import json, os, fcntl, time
+import json, os, fcntl
 import subprocess
 import hashlib
 
-from functools import reduce
+#from corebridge.aicorebridge import AICoreModule
 
 # %% ../nbs/02_rscriptbridge.ipynb 7
 def get_asset_path(script_name, assets_dir:str): 

diff --git a/nbs/00_core.ipynb b/nbs/00_core.ipynb
@@ -36,7 +36,7 @@
    "outputs": [],
    "source": [
     "#| export\n",
-    "import typing\n",
+    "import typing, logging\n",
     "import os\n",
     "import numpy as np\n",
     "import pandas as pd"
@@ -49,8 +49,9 @@
    "outputs": [],
    "source": [
     "#| export\n",
+    "syslog = logging.getLogger(__name__)\n",
     "try:\n",
-    "    print(f\"Loading {__name__} from {__file__}\")\n",
+    "    syslog.info()(f\"Loading {__name__} from {__file__}\")\n",
     "except:\n",
     "    pass"
    ]
@@ -287,7 +288,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "#| export\n",
+    "#| exports\n",
     "def timeseries_dataframe_from_datadict(\n",
     "        data:dict, \n",
     "        timecolumns=None,\n",
@@ -302,18 +303,28 @@
     "    Returns:\n",
     "        - df: A pandas DataFrame with a DatetimeIndex representing the converted data.\n",
     "    \"\"\"\n",
+    "\n",
     "    orient = recordformat.lower()\n",
     "    assert orient in ['records', 'table', 'split', 'index', 'tight']\n",
-    "    \n",
+    "    assert timecolumns, 'No time columns specified'\n",
+    "\n",
     "    if orient == 'records':\n",
+    "        # data is a structured ndarray, sequence of tuples or dicts, or DataFrame\n",
     "        df = pd.DataFrame.from_records(data)\n",
-    "        time_column = [C for C in df.columns if C in timecolumns][0]\n",
+    "        time_columns_in_df = [C for C in df.columns if C in timecolumns]\n",
+    "        if not  time_columns_in_df:\n",
+    "            syslog.error(f\"No  column in records {df.columns} matches specification in time columns {timecolumns}, assuming first column is time\")\n",
+    "            time_column = df.columns[0]\n",
+    "        else:\n",
+    "            time_column = time_columns_in_df[0]\n",
     "\n",
     "    elif orient == 'table':\n",
+    "        # data is in pandas table format\n",
     "        time_column = data['schema']['primaryKey'][0]\n",
     "        df = pd.DataFrame.from_dict(data['data']).set_index(data['schema']['primaryKey'])\n",
     "        df.index.name = 'time'\n",
     "    else:\n",
+    "        # data  is formatted according to 'orient' parameter (pandas)\n",
     "        df = pd.DataFrame.from_dict(data, orient=orient)\n",
     "        time_column = df.index.name\n",
     "\n",

diff --git a/nbs/01_aicorebridge.ipynb b/nbs/01_aicorebridge.ipynb
@@ -44,7 +44,15 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Loading corebridge.core from /home/fenke/repos/corebridge/corebridge/core.py\n"
+     ]
+    }
+   ],
    "source": [
     "#| export\n",
     "\n",
@@ -843,7 +851,7 @@
       "]\n",
       "Result Message\n",
       " [\n",
-      "  \"Startup time: 2024-07-08T13:46:30.324776+00:00\",\n",
+      "  \"Startup time: 2024-07-17T09:51:57.026890+00:00\",\n",
       "  \"test_function((data: pandas.core.frame.DataFrame, anumber: float | numpy.ndarray = 0))\",\n",
       "  \"init_args: (1, 2), init_kwargs: {'num_1': 3, 'num_2': 4, 'assets_dir': None, 'save_dir': '/home/fenke/repos/corebridge/nbs/cache'}\",\n",
       "  \"lastSeen: False, recordformat: records, timezone: Europe/Amsterdam\",\n",
@@ -947,7 +955,15 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Loading corebridge.aicorebridge 0.2.16 from /home/fenke/repos/corebridge/corebridge/aicorebridge.py\n"
+     ]
+    }
+   ],
    "source": [
     "\n",
     "from corebridge.aicorebridge import AICoreModule\n"
@@ -994,7 +1010,7 @@
       "]\n",
       "Result Message\n",
       " [\n",
-      "  \"Startup time: 2024-07-08T13:46:30.367098+00:00\",\n",
+      "  \"Startup time: 2024-07-17T09:51:57.093724+00:00\",\n",
       "  \"test_function((data: pandas.core.frame.DataFrame, anumber: float | numpy.ndarray = 0))\",\n",
       "  \"init_args: (1, 2), init_kwargs: {'num_1': 3, 'num_2': 4, 'assets_dir': None, 'save_dir': '/home/fenke/repos/corebridge/nbs/cache'}\",\n",
       "  \"lastSeen: False, recordformat: records, timezone: Europe/Amsterdam\",\n",

diff --git a/nbs/02_rscriptbridge.ipynb b/nbs/02_rscriptbridge.ipynb
@@ -35,7 +35,7 @@
    "outputs": [],
    "source": [
     "from nbdev.showdoc import *\n",
-    "import addroot"
+    "#import addroot"
    ]
   },
   {
@@ -46,11 +46,11 @@
    "source": [
     "#| export\n",
     "\n",
-    "import json, os, fcntl, time\n",
+    "import json, os, fcntl\n",
     "import subprocess\n",
     "import hashlib\n",
     "\n",
-    "from functools import reduce"
+    "#from corebridge.aicorebridge import AICoreModule"
    ]
   },
   {
@@ -93,6 +93,9 @@
    "outputs": [],
    "source": [
     "#| hide\n",
+    "\n",
+    "from functools import reduce\n",
+    "\n",
     "from IPython.display import display\n",
     "from IPython.display import Markdown\n",
     "\n",
@@ -186,7 +189,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Sapflow prediction scripts\n"
+    "## Example: sapflow prediction scripts\n"
    ]
   },
   {
@@ -509,7 +512,7 @@
     {
      "data": {
       "text/plain": [
-       "['lubridate', 'zoo', 'stringr', 'mgcv']"
+       "['lubridate', 'stringr', 'mgcv', 'zoo']"
       ]
      },
      "execution_count": null,
@@ -592,6 +595,12 @@
       "    date, intersect, setdiff, union\n",
       "\n",
       "\n",
+      "Library stringr already installed\n",
+      "\n",
+      "Library mgcv already installed\n",
+      "Loading required package: nlme\n",
+      "This is mgcv 1.9-1. For overview type 'help(\"mgcv-package\")'.\n",
+      "\n",
       "Library zoo already installed\n",
       "\n",
       "Attaching package: ‘zoo’\n",
@@ -600,12 +609,6 @@
       "\n",
       "    as.Date, as.Date.numeric\n",
       "\n",
-      "\n",
-      "Library stringr already installed\n",
-      "\n",
-      "Library mgcv already installed\n",
-      "Loading required package: nlme\n",
-      "This is mgcv 1.9-1. For overview type 'help(\"mgcv-package\")'.\n",
       "\n"
      ]
     }
@@ -1112,7 +1115,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### Putting it together\n",
+    "## Putting it together\n",
     "\n",
     "We need to run a script when either any of it's inputs have changed or any \n",
     "of it's outputs do not exist. Return True if a follow-up script should be \n",
@@ -1202,6 +1205,23 @@
     "run_script(data_file_flow[0], assets_dir, save_dir) and run_script(data_file_flow[1], assets_dir, save_dir) and run_script(data_file_flow[2], assets_dir, save_dir)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### AICore module class"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class AICoreRModule(AICoreModule):\n",
+    "    pass"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},

diff --git a/settings.ini b/settings.ini
@@ -5,7 +5,7 @@
 ### Python library ###
 repo = corebridge
 lib_name = %(repo)s
-version = 0.2.15
+version = 0.2.16
 min_python = 3.10
 license = apache2
 black_formatting = False