xmacis_violin_plot.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/tjturnage/climate/blob/main/xmacis_violin_plot.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<font size=\"+2\" color=\"blue\">xmACIS2 violin plots</font>  \n",
        "<font color=\"blue\">*v1.01 - updated 2/18/2024*</font>\n",
        "<br />  \n",
        "*T.J. Turnage - NWS Grand Rapids, MI -- thomas.turnage@noaa.gov*  \n",
        "<br />\n",
        "<hr>\n",
        "<br/>\n",
        "\n",
        "This leverages <a href=\"https://www.rcc-acis.org/docs_webservices.html\" target=\"_blank\">ACIS Web Services</a>\n",
        "<br />\n",
        "<br />\n",
        "It is recommended that you execute the following menu command to force a hard reset of the environment, which should help prevent execution errors: <br>\n",
        "* <font color=\"blue\">*Runtime > Disconnect and delete runtime*</font>  \n",
        "\n",
        "<br />\n",
        "\n",
        "<font size=\"+2\" color=\"blue\">*Pending issues*</font>\n",
        "* There are duplicate displays of the dashboard interface, not a deal-breaker just something to be aware of\n",
        "\n",
        "<br />\n",
        "\n",
        "<hr>"
      ],
      "metadata": {
        "id": "nTfrR-Pu6hOb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# @title <b>Step 1.</b> <i><font color=\"blue\">Install the packages you need.</font></i>\n",
        "!pip install jupyter-dash &> /dev/null\n",
        "!pip install dash-bootstrap-components &> /dev/null\n",
        "import json\n",
        "import requests\n",
        "from datetime import datetime\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from jupyter_dash import JupyterDash\n",
        "from dash import dcc, html, Output, Input, no_update\n",
        "import dash_bootstrap_components as dbc\n",
        "import plotly.graph_objects as go\n",
        "import plotly.express as px\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "metadata": {
        "id": "TnXR17_asPuC",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title <b>Step 2.</b> <i><font color=\"blue\">Determine a station you want to get data for<br>Long term threaded datasets end with \"thr\".\n",
        "meta_url = \"http://data.rcc-acis.org/StnMeta\"\n",
        "\n",
        "station_name = \"GRRthr\" #@param [\"GRRthr\"] {allow-input: true}\n",
        "\n",
        "try:\n",
        "    params = {\"sids\": station_name, \"elems\": \"maxt\", \"meta\": [\"name\",\"valid_daterange\"]}\n",
        "    req = requests.post(meta_url,json=params)\n",
        "    data = req.json()\n",
        "    meta = data['meta'][0]\n",
        "    name = meta['name']\n",
        "    print(f'The default name of the station is: {name}.')\n",
        "except Exception:\n",
        "  print(\"Station Not Found! Check your entry and try again.\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oXSM1nSTUmsq",
        "outputId": "cfd1f8ba-d50b-4c54-ff79-2c87c188ee59",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The default name of the station is: Grand Rapids Area.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title <b>Step 3.</b> <i><font color=\"blue\">If you don't like the default station name above, type what you'd rather have displayed. <br>Otherwise ensure field below is blank.</font></i>\n",
        "new_name= \"Grand Rapids\" #@param {type:\"string\"}"
      ],
      "metadata": {
        "id": "WzFXFCtgwAGm",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title <b>Step 4.</b> <i><font color=\"blue\">Retrieve the climate dataset</i></font>\n",
        "mon = {1:'Jan', 2:'Feb', 3:'Mar', 4:'Apr', 5:'May', 6:'Jun', 7:'Jul', 8:'Aug', 9:'Sep', 10:'Oct', 11:'Nov', 12:'Dec'}\n",
        "months = list(np.arange(1,13))\n",
        "dates = list(np.arange(1,32))\n",
        "\n",
        "parms_url = \"http://data.rcc-acis.org/StnData?params=\"\n",
        "\n",
        "class XMACIS():\n",
        "    def __init__(self,station):\n",
        "\n",
        "        self.station = station\n",
        "\n",
        "        #grab metadata related to station's Period Of Record (POR)\n",
        "        self.por_metadata = self.get_por_metadata()\n",
        "        self.df = self.get_all_por_data()\n",
        "        self.melted = pd.melt(self.df, id_vars=['year','month','day'], value_vars=['mint', 'maxt', 'avgt', 'maxt_n','mint_n','avgt_n'], value_name='temperature')\n",
        "        self.melted = self.melted.astype({\"variable\": 'string'}, errors='ignore')\n",
        "\n",
        "\n",
        "    def get_por_metadata(self):\n",
        "        \"\"\"\n",
        "        Returns list of strings: ['YYYY-mm-dd','YYYY-mm-dd']\n",
        "        Refers to first/last dates ofperiod of record\n",
        "        \"\"\"\n",
        "        params = {\"sids\": self.station, \"elems\": \"maxt\", \"meta\": [\"name\",\"valid_daterange\"]}\n",
        "        req = requests.post(meta_url,json=params)\n",
        "        data = req.json()\n",
        "        meta = data['meta'][0]\n",
        "        self.full_name = meta['name']\n",
        "        try:\n",
        "          self.full_name = new_name\n",
        "        except:\n",
        "          pass\n",
        "        self.por_start_date_str = meta['valid_daterange'][0][0]\n",
        "        self.por_end_date_str = meta['valid_daterange'][0][1]\n",
        "        self.por_start_date_obj = datetime.strptime(self.por_start_date_str, '%Y-%m-%d')\n",
        "        self.por_end_date_obj = datetime.strptime(self.por_end_date_str, '%Y-%m-%d')\n",
        "        self.por_start_year = self.por_start_date_obj.year\n",
        "        self.por_end_year = self.por_end_date_obj.year\n",
        "        self.por_year_range = self.por_end_year - self.por_start_year + 1\n",
        "        self.por_year_index = pd.date_range(start=f'1/1/{self.por_start_year}', periods=self.por_year_range, freq='Y')\n",
        "        self.por_label = f'{self.por_start_year} - {self.por_end_year}'\n",
        "        return\n",
        "\n",
        "    def get_all_por_data(self):\n",
        "        parms = {\n",
        "        \"sid\":self.station,\n",
        "        \"sdate\":self.por_start_date_str,\n",
        "        \"edate\":self.por_end_date_str,\n",
        "        \"elems\":[\n",
        "            {\"name\":\"maxt\",\"interval\":\"dly\",\"duration\":\"dly\"},\n",
        "            {\"name\":\"mint\",\"interval\":\"dly\",\"duration\":\"dly\"},\n",
        "            {\"name\":\"avgt\",\"interval\":\"dly\",\"duration\":\"dly\"},\n",
        "            {\"name\":\"maxt\",\"interval\":\"dly\",\"duration\":\"dly\",\"normal\":\"1\"},\n",
        "            {\"name\":\"mint\",\"interval\":\"dly\",\"duration\":\"dly\",\"normal\":\"1\"},\n",
        "            {\"name\":\"avgt\",\"interval\":\"dly\",\"duration\":\"dly\",\"normal\":\"1\"},\n",
        "            ]}\n",
        "\n",
        "        req = requests.post(parms_url,json=parms)\n",
        "        data = req.json()\n",
        "        df_init = pd.json_normalize(data)\n",
        "        df_dp = df_init[\"data\"].apply(pd.Series).T\n",
        "        df = df_dp[0].apply(pd.Series)\n",
        "        df.columns = ['dts','maxt','mint','avgt','maxt_n','mint_n','avgt_n']\n",
        "        df['date'] = pd.to_datetime(df['dts'], errors='coerce')\n",
        "        #df['date'] = pd.DatetimeIndex(df['dts'])\n",
        "        df.set_index('date', inplace=True)\n",
        "        df['year'] = df.index.year\n",
        "        df['month'] = df.index.month\n",
        "        df['day'] = df.index.day\n",
        "        df = df.astype({\"dts\":'string', \"year\":'int', \"month\":'int', \"day\":'int', \"maxt\":'int', \"mint\":'int', \"avgt\":'float', \"maxt_n\":'float', \"mint_n\":'float', \"avgt_n\":'float'}, errors='ignore')\n",
        "        return df\n",
        "\n",
        "\n",
        "test = XMACIS(station_name)\n",
        "\n",
        "dff = test.melted\n",
        "por_start = test.por_start_year\n",
        "por_end = test.por_end_year\n",
        "yrs = list(range(por_start,por_end+1))\n",
        "dff = dff.astype({\"variable\": 'string'}, errors='ignore')\n",
        "dff['mon'] = dff['month'].map(mon)\n",
        "\n",
        "D = test.df\n",
        "# remove leap days ... too much trouble\n",
        "D = D[~((D.index.month == 2) & (D.index.day == 29))]\n",
        "#output = test.df.to_csv(index=False,header=True)\n",
        "#with open('climate-data.csv', 'w') as outfile:\n",
        "#  for lines in output:\n",
        "#    outfile.write(lines)"
      ],
      "metadata": {
        "id": "VV4Nj8_95AD9",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oAV519rsrpCk",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@title <b>Step 5.</b> <i><font color=\"blue\">Run code to plot data. Once it completes, scroll down to see interactive graph.<br>(this might take a while to render at first)</font></i>\n",
        "\n",
        "\n",
        "\n",
        "app = JupyterDash(__name__, external_stylesheets=[dbc.themes.CYBORG])\n",
        "\n",
        "app.layout = html.Div([\n",
        "    dbc.Row(html.H6(' ')),\n",
        "    dbc.Row([\n",
        "        dbc.Col([\n",
        "            html.H6('Graph Type'),\n",
        "            dcc.Dropdown(['violin'], value='violin', id='graph_type',searchable=False)\n",
        "        ]),\n",
        "        dbc.Col([\n",
        "            html.H6('Start Date'),\n",
        "            dcc.Dropdown(yrs, value=por_start, id='base_start',searchable=False)\n",
        "        ]),\n",
        "        dbc.Col([\n",
        "            html.H6('Select end year'),\n",
        "            dcc.Dropdown(yrs, value=por_end, id='base_end',searchable=False)\n",
        "        ]),\n",
        "        dbc.Col([\n",
        "            html.H6('Select element to investigate'),\n",
        "            dcc.Dropdown(\n",
        "                options=[\n",
        "                    {'label': 'Min Temp', 'value': 'mint'},\n",
        "                    {'label': 'Max Temp', 'value': 'maxt'},\n",
        "                    {'label': 'Avg Temp', 'value': 'avgt'}\n",
        "                    ],\n",
        "                    value='maxt',id=\"element\", searchable=False)\n",
        "        ]),\n",
        "        dbc.Col([\n",
        "            html.H6('Year to compare'),\n",
        "            dcc.Dropdown(yrs, value=2015, id='single_year',searchable=False)\n",
        "        ])\n",
        "    ]),\n",
        "    dbc.Row(html.H6(' ')),\n",
        "    dbc.Row(dcc.Graph(id=\"graphing\"))\n",
        "])\n",
        "\n",
        "\n",
        "@app.callback(\n",
        "    Output(\"graphing\", \"figure\"),\n",
        "    Input(\"graph_type\", \"value\"),\n",
        "    Input(\"base_start\", \"value\"),\n",
        "    Input(\"base_end\", \"value\"),\n",
        "    Input(\"element\", \"value\"),\n",
        "    Input(\"single_year\", \"value\"))\n",
        "def update_violin_chart(graph_type,base_start,base_end,element,single_year):\n",
        "  df = dff[(dff['variable'] == element)]\n",
        "  nice_name = 'Minimum Temperature'\n",
        "  col = 'blue'\n",
        "  if element == 'maxt':\n",
        "    nice_name = 'Maximum Temperature'\n",
        "    col = 'red'\n",
        "  if element == 'avgt':\n",
        "    nice_name = 'Average Temperature'\n",
        "    col = 'orange'\n",
        "  df_base = df[(df['year'] >= base_start) & (df['year'] <= base_end)]\n",
        "  df_year = df[df['year'] == single_year]\n",
        "\n",
        "  fig = go.Figure()\n",
        "  fig.add_trace(go.Violin(x=df_base['mon'], y=df_base['temperature'],\n",
        "                          legendgroup='Yes', scalegroup='Yes', name=f'{base_start}-{base_end}',\n",
        "                          side='negative',\n",
        "                          line_color='gray')\n",
        "              )\n",
        "  fig.add_trace(go.Violin(x=df_year['mon'],\n",
        "                          y=df_year['temperature'],\n",
        "                          legendgroup='No', scalegroup='No', name=f'{single_year}',\n",
        "                          side='positive',\n",
        "                          line_color=col)\n",
        "              )\n",
        "  fig.update_traces(meanline_visible=True)\n",
        "  fig.update_layout(violingap=0, violinmode='overlay')\n",
        "  r = [-30,120]\n",
        "  fig.update_yaxes(range = r)\n",
        "  fig.update_xaxes(categoryorder='array', categoryarray= ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'])\n",
        "  fig.update_layout(autotypenumbers='convert types')\n",
        "  fig.update_layout(title=f'{test.full_name} - {nice_name}',\n",
        "                    legend_title='',\n",
        "                    yaxis_title='Temperature  (F)')\n",
        "  fig.update_layout(margin=dict(l=20, r=10, t=60, b=20))\n",
        "\n",
        "  return fig\n",
        "\n",
        "app.run_server(mode='inline', port=8030)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# @title <b>Ignore code below</b>\n",
        "months = list(np.arange(1,13))\n",
        "dates = list(np.arange(1,32))\n",
        "\n",
        "\n",
        "#app.layout = html.Div([\n",
        "#    dbc.Row(html.H6(' ')),\n",
        "#    dbc.Row([\n",
        "#        dbc.Col([\n",
        "#            html.H6('Select year'),\n",
        "#            dcc.Dropdown(yrs, value=2015, id='year',searchable=False)\n",
        "#        ]),\n",
        "#        dbc.Col([\n",
        "#            html.H6('Select month'),\n",
        "#            dcc.Dropdown(months, value=6, id='month',searchable=False)\n",
        "#        ]),\n",
        "#        dbc.Col([\n",
        "#            html.H6('Select element'),\n",
        "#            dcc.Dropdown(\n",
        "#                options=[\n",
        "#                    {'label': 'Min Temp', 'value': 'mint'},\n",
        "#                    {'label': 'Max Temp', 'value': 'maxt'},\n",
        "#                    {'label': 'Avg Temp', 'value': 'avgt'}#,\n",
        "#                    ],\n",
        "#                    value='maxt',id=\"heatmap_element\", searchable=False)\n",
        "#        ]),\n",
        "#    ]),\n",
        "#    dbc.Row(html.H6(' ')),\n",
        "#    dbc.Row(dcc.Graph(id=\"heatmap\"))\n",
        "#])\n",
        "\n",
        "\n",
        "#@app.callback(\n",
        "#    Output(\"graphing\", \"figure\"),\n",
        "#    Input(\"year\", \"value\"),\n",
        "#    #Input(\"month\", \"value\"),\n",
        "#    Input(\"heatmap_element\", \"value\"))\n",
        "year=2014\n",
        "month = 2\n",
        "element = 'maxt'\n",
        "def update_heatmap(year,month,element):\n",
        "  df = dff[(dff['variable'] == element)]\n",
        "  nice_name = 'Minimum Temperature'\n",
        "  col = 'blue'\n",
        "  if element == 'maxt':\n",
        "    nice_name = 'Maximum Temperature'\n",
        "    col = 'red'\n",
        "  if element == 'avgt':\n",
        "    nice_name = 'Average Temperature'\n",
        "    col = 'orange'\n",
        "  df_heatmap = df[(df['year'] >= year) & (df.index.month == month)]\n",
        "\n",
        "\n",
        "  #fig = go.Figure(data=go.Heatmap(\n",
        "          #x = months,\n",
        "          #y = dates,\n",
        "          #z = df_heatmap[element],\n",
        "          #type = 'heatmap',\n",
        "          #colorscale = 'Viridis'))\n",
        "\n",
        "  fig = go.Figure(data=go.Heatmap(\n",
        "          x = [1,2,3],\n",
        "          y = [4,5,6],\n",
        "          z = [[9,8,7],\n",
        "               [6,5,4],\n",
        "               [3,2,1]],\n",
        "          type = 'heatmap',\n",
        "          colorscale = 'Viridis'))\n",
        "\n",
        "\n",
        "  #fig.update_traces(meanline_visible=True)\n",
        "  #fig.update_layout(violingap=0, violinmode='overlay')\n",
        "  #r = [-30,120]\n",
        "  #fig.update_yaxes(range = r)\n",
        "  #fig.update_layout(autotypenumbers='convert types')\n",
        "  #fig.update_layout(title=f'{test.full_name} - {nice_name}',\n",
        "  #                  legend_title='',\n",
        "  #                  yaxis_title='Temperature  (F)')\n",
        "  #fig.update_layout(margin=dict(l=20, r=10, t=60, b=20))\n",
        "  return fig\n",
        "\n",
        "#app.run_server(mode='inline', port=8030)\n",
        "import plotly.express as px\n",
        "import plotly.graph_objects as go\n",
        "from datetime import datetime,timedelta\n",
        "\n",
        "day_num = []\n",
        "doy_strings = []\n",
        "for doy in range(0,365):\n",
        "  base = datetime(1999,1,1)\n",
        "  daymonth_datetime = base + timedelta(days=int(doy))\n",
        "  doy_strings.append(daymonth_datetime.strftime('%b-%d'))\n",
        "\n",
        "x_label = []\n",
        "for ii, d in enumerate(doy_strings):\n",
        "  if '01' in d:\n",
        "    day_num.append(ii)\n",
        "    x_label.append(d)\n",
        "\n",
        "df = test.df\n",
        "df_base = df['2010-01-01' : '2020-12-31']\n",
        "#print(df_base)\n",
        "\n",
        "\n",
        "\n",
        "fig1 = px.scatter(x=df.index.dayofyear, y=df['maxt'], color=df.index.year)\n",
        "fig2 = px.scatter(x=df_base.index.dayofyear, y=df_base['maxt'])\n",
        "fig2.update_traces(marker=dict(color='black'))\n",
        "fig3 = go.Figure(data=fig1.data + fig2.data)\n",
        "fig3.update_layout(autotypenumbers='convert types')\n",
        "fig3.show()\n",
        "#fig.add_scatter(x=df_base.index.dayofyear, y=df_base['maxt_n'], linestyle=':', color='r', linewidth=3, alpha=0.75) # plot normals\n",
        "\n"
      ],
      "metadata": {
        "id": "2XeQRVOVsOK_",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}