data/issues.ipynb

   1 {
   2  "cells": [
   3   {
   4    "cell_type": "code",
   5    "execution_count": null,
   6    "metadata": {},
   7    "outputs": [
   8     {
   9      "ename": "",
  10      "evalue": "",
  11      "output_type": "error",
  12      "traceback": [
  13       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
  14       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
  15       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
  16      ]
  17     }
  18    ],
  19    "source": [
  20     "import platform\n",
  21     "import matplotlib.pyplot as plt\n",
  22     "print(platform.python_version())"
  23    ]
  24   },
  25   {
  26    "cell_type": "markdown",
  27    "metadata": {},
  28    "source": []
  29   },
  30   {
  31    "cell_type": "markdown",
  32    "metadata": {},
  33    "source": [
  34     "# Todo List\n",
  35     "\n",
  36     "- [x] add calculated year to the metadata\n",
  37     "- [ ] make one chart per year\n",
  38     "- [x] make sure all the data files are getting loaded"
  39    ]
  40   },
  41   {
  42    "cell_type": "code",
  43    "execution_count": null,
  44    "metadata": {},
  45    "outputs": [
  46     {
  47      "ename": "",
  48      "evalue": "",
  49      "output_type": "error",
  50      "traceback": [
  51       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
  52       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
  53       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
  54      ]
  55     }
  56    ],
  57    "source": [
  58     "# Get ye data\n",
  59     "import glob\n",
  60     "import os\n",
  61     "\n",
  62     "data_files = glob.glob(os.path.expanduser('~\\\\Data\\\\*.json'))\n",
  63     "\n",
  64     "for file in data_files:\n",
  65     "\tprint(f\"{file}\", os.path.getsize(file))\n",
  66     "\n"
  67    ]
  68   },
  69   {
  70    "cell_type": "code",
  71    "execution_count": null,
  72    "metadata": {},
  73    "outputs": [
  74     {
  75      "ename": "",
  76      "evalue": "",
  77      "output_type": "error",
  78      "traceback": [
  79       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
  80       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
  81       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
  82      ]
  83     }
  84    ],
  85    "source": [
  86     "import pandas as pd\n",
  87     "\n",
  88     "data_frames = []\n",
  89     "for file in data_files:\n",
  90     "\n",
  91     "\tf = open(file, encoding='utf-8')\n",
  92     "\tdf_temp = pd.read_json(f)\n",
  93     "\tdata_frames.append(df_temp)\n",
  94     "\n",
  95     "df = pd.concat(data_frames, ignore_index=True)"
  96    ]
  97   },
  98   {
  99    "cell_type": "code",
 100    "execution_count": null,
 101    "metadata": {},
 102    "outputs": [
 103     {
 104      "ename": "",
 105      "evalue": "",
 106      "output_type": "error",
 107      "traceback": [
 108       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 109       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 110       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 111      ]
 112     }
 113    ],
 114    "source": [
 115     "import numpy as np\n",
 116     "df['count'] = np.zeros(len(df))\n",
 117     "\n"
 118    ]
 119   },
 120   {
 121    "cell_type": "markdown",
 122    "metadata": {},
 123    "source": [
 124     "#### Pie Chart"
 125    ]
 126   },
 127   {
 128    "cell_type": "code",
 129    "execution_count": null,
 130    "metadata": {},
 131    "outputs": [
 132     {
 133      "ename": "",
 134      "evalue": "",
 135      "output_type": "error",
 136      "traceback": [
 137       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 138       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 139       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 140      ]
 141     }
 142    ],
 143    "source": [
 144     "\n",
 145     "dfp = df\n",
 146     "# dfp['mile'] = [m.get('title') if m else None for m in dfp.milestone]\n",
 147     "dfp['mile'] = [m.get('title') if hasattr(m, 'get') else None for m in dfp['milestone']]\n"
 148    ]
 149   },
 150   {
 151    "cell_type": "code",
 152    "execution_count": null,
 153    "metadata": {},
 154    "outputs": [
 155     {
 156      "ename": "",
 157      "evalue": "",
 158      "output_type": "error",
 159      "traceback": [
 160       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 161       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 162       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 163      ]
 164     }
 165    ],
 166    "source": [
 167     "closed_dates = pd.to_datetime(df['closedAt'], format='ISO8601')\n",
 168     "dfp['year'] = closed_dates.dt.year\n",
 169     "\n",
 170     "dfp.drop(columns=['closedAt','milestone', 'url'], inplace=True)\n",
 171     "print(dfp)"
 172    ]
 173   },
 174   {
 175    "cell_type": "code",
 176    "execution_count": null,
 177    "metadata": {},
 178    "outputs": [
 179     {
 180      "ename": "",
 181      "evalue": "",
 182      "output_type": "error",
 183      "traceback": [
 184       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 185       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 186       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 187      ]
 188     }
 189    ],
 190    "source": [
 191     "\n",
 192     "mile_counts = dfp.groupby('mile')['title'].nunique()\n",
 193     "dfp['count'] = dfp['mile'].map(mile_counts)\n",
 194     "# dfp.drop(columns=['title'], inplace=True)\n",
 195     "dfp['mile'] = df['mile'].drop_duplicates()\n",
 196     "dfp.dropna(inplace=True)\n",
 197     "dfp = dfp[dfp['count'] > 5]\n",
 198     "print(dfp)"
 199    ]
 200   },
 201   {
 202    "cell_type": "code",
 203    "execution_count": null,
 204    "metadata": {},
 205    "outputs": [
 206     {
 207      "ename": "",
 208      "evalue": "",
 209      "output_type": "error",
 210      "traceback": [
 211       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 212       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 213       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 214      ]
 215     }
 216    ],
 217    "source": [
 218     "mile_counts = df.groupby('mile')['count'].sum()\n",
 219     "mile_counts = mile_counts.loc[lambda x: x > 5]\n",
 220     "\n",
 221     "\n",
 222     "print(mile_counts)"
 223    ]
 224   },
 225   {
 226    "cell_type": "code",
 227    "execution_count": null,
 228    "metadata": {},
 229    "outputs": [
 230     {
 231      "ename": "",
 232      "evalue": "",
 233      "output_type": "error",
 234      "traceback": [
 235       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 236       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 237       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 238      ]
 239     }
 240    ],
 241    "source": [
 242     "\n",
 243     "plt.pie(mile_counts, labels=mile_counts.index, autopct='%0.1f%%')\n",
 244     "plt.title(f\"Tasks by Milestone\")\n",
 245     "plt.axis('equal')\n",
 246     "plt.show()"
 247    ]
 248   },
 249   {
 250    "cell_type": "code",
 251    "execution_count": null,
 252    "metadata": {},
 253    "outputs": [
 254     {
 255      "ename": "",
 256      "evalue": "",
 257      "output_type": "error",
 258      "traceback": [
 259       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 260       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 261       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 262      ]
 263     }
 264    ],
 265    "source": [
 266     "type(mile_counts)"
 267    ]
 268   },
 269   {
 270    "cell_type": "markdown",
 271    "metadata": {},
 272    "source": [
 273     "#### Plot of closed issues by month and milestone for one project\n"
 274    ]
 275   },
 276   {
 277    "cell_type": "code",
 278    "execution_count": null,
 279    "metadata": {},
 280    "outputs": [
 281     {
 282      "ename": "",
 283      "evalue": "",
 284      "output_type": "error",
 285      "traceback": [
 286       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 287       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 288       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 289      ]
 290     }
 291    ],
 292    "source": [
 293     "closed_months = pd.to_datetime(df['closedAt'], format='ISO8601')\n",
 294     "df['closed_month'] = closed_months.dt.month\n",
 295     "df.drop(columns=['closedAt'], inplace=True)"
 296    ]
 297   },
 298   {
 299    "cell_type": "code",
 300    "execution_count": null,
 301    "metadata": {},
 302    "outputs": [
 303     {
 304      "ename": "",
 305      "evalue": "",
 306      "output_type": "error",
 307      "traceback": [
 308       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 309       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 310       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 311      ]
 312     }
 313    ],
 314    "source": [
 315     "df['mile'] = [d.get('title') if d else None for d in df.milestone]"
 316    ]
 317   },
 318   {
 319    "cell_type": "code",
 320    "execution_count": null,
 321    "metadata": {},
 322    "outputs": [
 323     {
 324      "ename": "",
 325      "evalue": "",
 326      "output_type": "error",
 327      "traceback": [
 328       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 329       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 330       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 331      ]
 332     }
 333    ],
 334    "source": [
 335     "\n",
 336     "dfg = df.groupby(['mile', 'closed_month']).count().reset_index()\n",
 337     "dfg.drop(columns=['milestone', 'title', 'url'], inplace=True)\n"
 338    ]
 339   },
 340   {
 341    "cell_type": "code",
 342    "execution_count": null,
 343    "metadata": {},
 344    "outputs": [
 345     {
 346      "ename": "",
 347      "evalue": "",
 348      "output_type": "error",
 349      "traceback": [
 350       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 351       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 352       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 353      ]
 354     }
 355    ],
 356    "source": [
 357     "exclude_from_chart = [\n",
 358     "\t'Job Aid Ideas']\n",
 359     "\n",
 360     "for ex in exclude_from_chart:\n",
 361     "\tdfg = dfg[dfg['mile'] != ex]\n"
 362    ]
 363   },
 364   {
 365    "cell_type": "code",
 366    "execution_count": null,
 367    "metadata": {},
 368    "outputs": [
 369     {
 370      "ename": "",
 371      "evalue": "",
 372      "output_type": "error",
 373      "traceback": [
 374       "\u001b[1;31mRunning cells with '/bin/python3' requires the ipykernel package.\n",
 375       "\u001b[1;31mRun the following command to install 'ipykernel' into the Python environment. \n",
 376       "\u001b[1;31mCommand: '/bin/python3 -m pip install ipykernel -U --user --force-reinstall'"
 377      ]
 378     }
 379    ],
 380    "source": [
 381     "import seaborn as sns\n",
 382     "import matplotlib.pyplot as plt\n",
 383     "sns.set_theme()"
 384    ]
 385   },
 386   {
 387    "cell_type": "code",
 388    "execution_count": null,
 389    "metadata": {},
 390    "outputs": [],
 391    "source": [
 392     "plt.figure(figsize=(10, 6))\n",
 393     "sns.pie(x='closed_month', y='count', hue='mile', data=dfg)\n",
 394     "plt.pie(dfg, )\n",
 395     "plt.xlabel('Closed Month')\n",
 396     "plt.ylabel('ClosedAt')\n",
 397     "plt.title('Stacked Line Chart: Milestone Effort by Issue Close Date')\n",
 398     "plt.xticks(rotation=45)\n",
 399     "plt.legend(title='Mile', loc='upper left')\n",
 400     "plt.show()"
 401    ]
 402   }
 403  ],
 404  "metadata": {
 405   "kernelspec": {
 406    "display_name": "Python 3",
 407    "language": "python",
 408    "name": "python3"
 409   },
 410   "language_info": {
 411    "codemirror_mode": {
 412     "name": "ipython",
 413     "version": 3
 414    },
 415    "file_extension": ".py",
 416    "mimetype": "text/x-python",
 417    "name": "python",
 418    "nbconvert_exporter": "python",
 419    "pygments_lexer": "ipython3",
 420    "version": "3.8.10"
 421   }
 422  },
 423  "nbformat": 4,
 424  "nbformat_minor": 2
 425 }