use gbif

h3_ibis.ipynb

@@ -1,177 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "ae26cf1d-f639-418a-8c16-365024a8182f",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import ibis\n",
-    "from ibis import _\n",
-    "\n",
-    "con = ibis.duckdb.connect(extensions=[\"h3\"])\n",
-    "\n",
-    "@ibis.udf.scalar.builtin\n",
-    "def h3_latlng_to_cell(lat: float, lng: float, zoom: int) -> int:\n",
-    "    ...\n",
-    "\n",
-    "@ibis.udf.scalar.builtin\n",
-    "def hex(array) -> str:\n",
-    "    ...\n",
-    "\n",
-    "\n",
-    "parquet = \"https://data.source.coop/cboettig/obis/obis_20240625.parquet\"\n",
-    "parquet = \"/home/rstudio/source.coop/cboettig/obis/obis_20240625.parquet\"\n",
-    "\n",
-    "obis = con.read_parquet(parquet)\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "ee13fff4-fa0d-4c5e-9546-9a788ccdec2e",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "%%time \n",
-    "\n",
-    "(obis\n",
-    " .mutate(h1 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 1)),\n",
-    "        h2 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 2)),\n",
-    "        h3 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 3)),\n",
-    "        h4 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 4)),\n",
-    "        h5 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 5)),\n",
-    "        h6 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 6)),\n",
-    "        h7 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 7)),\n",
-    "        h8 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 8)),\n",
-    "        h9 = hex(h3_latlng_to_cell(_.decimalLatitude, _.decimalLongitude, 9)))\n",
-    " .to_parquet(\"obis_h3.parquet\")\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "68b5b794-91cb-4b98-acb3-008b80516d3c",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "obis_h3 = con.read_parquet(\"obis_h3.parquet\")\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "e42f44fd-8668-41f0-b890-25918072f2c2",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "47e2ebce-f1cc-42af-a0f9-0ff155143bcc",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import pandas as pd\n",
-    "def get_h3point_df(resolution: float) -> pd.DataFrame:\n",
-    "    column = \"h\" + str(resolution)\n",
-    "    df = (obis_h3\n",
-    "            .rename(cell = column)\n",
-    "            .cell.value_counts()\n",
-    "            .mutate(v = _.cell_count.log())\n",
-    "            .mutate(normalized_values = _.v / _.v.max())\n",
-    "            .to_pandas()\n",
-    "           )\n",
-    "    return df\n",
-    "\n",
-    "\n",
-    "df = get_h3point_df(4)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "7c717c3b-ec75-4ffa-857f-bfbe8595847e",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "from matplotlib import cm\n",
-    "def viridis_color(x)  -> str:\n",
-    "    rgb_array = np.round( cm.viridis(x) * 255 ).astype(int).clip(0,255).tolist()\n",
-    "    return rgb_array\n",
-    "    \n",
-    "df['rgb'] = viridis_color(df.normalized_values)\n",
-    "df"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "854a8b05-4244-4f6d-b7a5-11cdd108e3d3",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import pydeck as pdk\n",
-    "from typing import List\n",
-    "\n",
-    "def get_coverage_layer(df: pd.DataFrame, line_color: List) -> pdk.Layer:\n",
-    "    return pdk.Layer(\n",
-    "        \"H3HexagonLayer\",\n",
-    "        df,\n",
-    "        get_hexagon=\"cell\",\n",
-    "        filled=True,\n",
-    "        auto_highlight=True,\n",
-    "        get_fill_color=\"rgb\",\n",
-    "      #  get_elevation=\"normalized_values\",\n",
-    "        pickable=True,\n",
-    "        extruded=False,\n",
-    "        line_width_min_pixels=1,\n",
-    "    )\n",
-    "\n",
-    "layer_coverage_1 = get_coverage_layer(df, line_color = [36, 191, 242])\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "dee7c1c1-1664-433c-9963-ae85f8d0ef85",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "pdk.Deck(map_provider='carto', \n",
-    "    map_style='light',\n",
-    "    initial_view_state=pdk.ViewState(\n",
-    "        latitude=30., longitude=0., zoom=3, height=400\n",
-    "    ),\n",
-    "    layers=[layer_coverage_1],\n",
-    ")"
-   ]
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.10.12"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}

main.py

@@ -9,45 +9,19 @@ import ibis
 from ibis import _
 
 image = Image.open('./favicon.png')
-st.set_page_config(page_title="H3: Simplifying the World's Map", page_icon=image)
-st.header("H3: Simplifying the World's Map", divider="rainbow")
+st.set_page_config(page_title="GBIF Observations in H3", page_icon=image)
+st.header("GBIF Observations in H3", divider="rainbow")
 
 # st.set_page_config(page_title="H3 in Streamlit", layout="wide")
 
-h3_parquet = "https://data.source.coop/cboettig/obis/obis_h3.parquet"
-h3_parquet = "/home/rstudio/source.coop/cboettig/obis/obis_h3.parquet"
+h3_parquet = "https://data.source.coop/cboettig/gbif/gbif_ca_h3.parquet"
+#h3_parquet = "/home/rstudio/source.coop/cboettig/obis/obis_h3.parquet"
 
 con = ibis.duckdb.connect(extensions=["h3"])
 obis_h3 = con.read_parquet(h3_parquet)
 
-st.subheader("What is H3")
+st.subheader("California total observations")
 
-col1, col2 = st.columns([0.7, 0.3])
-with col1:
-    st.markdown("H3 Discrete Global Grid is a way to divide the world into a grid of hexagonal cells of equal sizes, "
-            "each with a unique identifier (string or integer). It is a hierarchical grid, meaning that cells can be "
-            "aggregated into larger cells, and vice versa. This makes it very efficient for processing geospatial data.") 
-with col2:
-    st.image('https://viennadatasciencegroup.at/post/2019-11-21-h3spark/featured.png', 
-         width=180)
-    
-col1, col2, col3, col4 = st.columns(4)
-with col2:
-    st.write("**String**")
-    st.text('8c274daeb7a0bff')
-    st.text('8c2ab2d9294c5ff')
-    st.text('8c2ab2da36605ff')
-
-with col3:
-    st.write("**Integer**")
-    st.text('631195381387627519')
-    st.text('631255110006392319')
-    st.text('631255110288541183')
-
-st.markdown("The lowest resolution is 0, at which the world is divided into 122 hexagons. "
-            "The highest resolution is 15, at which the size of a hexagon is less than a square meter, "
-            "and the world can be divided into approximately 600 trillion hexagons."
-            " You can check different resolutions and play with hierarchy levels using the widget below.")
 
 # ------ Visualisation 1 ---------
 def viridis_color(x)  -> str:
@@ -76,20 +50,22 @@ def get_coverage_layer(df: pd.DataFrame) -> pdk.Layer:
         filled=True,
         auto_highlight=True,
         get_fill_color="rgb",
-        #get_elevation="normalized_values",
+        get_elevation="normalized_values",
+        elevation_scale=5000,
+        elevation_range=[0,1],
         pickable=True,
-        extruded=False,
+        extruded=True,
         line_width_min_pixels=1,
     )
 
-min_v_1, max_v_1, v_1, z_1, lon_1, lat_1 = ( 1, 9, 1, 1, 0.9982847947205775, 2.9819747220001886,)
+min_v_1, max_v_1, v_1, z_1, lon_1, lat_1 = ( 1, 9, 4, 4, -120, 40,)
 col1, col2 = st.columns([70, 30])
 with col1:
     h3_resolut_1 = st.slider(
         "H3 resolution", min_value=min_v_1, max_value=max_v_1, value=v_1)
 
 with col2:
-    levels_option = st.selectbox("Levels", ("One", "Two", "Three"))
+    levels_option = st.selectbox("Add filters", ("One", "Two", "Three"))
 
 df = get_h3point_df(h3_resolut_1)
 layer_coverage_1 = get_coverage_layer(df)
@@ -101,7 +77,7 @@ st.pydeck_chart(
         initial_view_state=pdk.ViewState(
             latitude=lat_1, longitude=lon_1, zoom=z_1, height=400
         ),
-        tooltip={"html": "<b>ID:</b> {cell}", "style": {"color": "white"}},
+        tooltip={"html": "<b>ID:</b> {cell_count}", "style": {"color": "white"}},
         layers=visible_layers_coverage_1,
     )
 )