Create container

README.md

@@ -6,3 +6,28 @@ Streamlit dashboard with analysis of cases and hospitalizations from Genebra.
 ```bash
 $ streamlit run dashboard.py --server.port 8502
 ```
+
+### Run with docker-compose :
+
+**Pre-requisites**
+* Docker installed and running
+* docker-compose installed
+
+It also works fine with podman and [podman-compose].(https://phoenixnap.com/kb/podman-compose#ftoc-heading-3)
+
+```bash
+$ docker-compose -f docker/docker-compose.yml up
+# or 
+$ podman-compose -f docker/docker-compose.yml up
+
+# When dependencies change and you need to force a rebuild
+$ docker-compose -f docker/docker-compose.yml up --build
+
+# When finished
+$ docker-compose down
+```
+
+You can see the app running at http://localhost:8501
+
+
+docker-compose -f docker/docker-compose.yaml --env-file .env build

SIR_forecast.py → app/SIR_forecast.py

@@ -17,8 +17,8 @@
 import matplotlib.pyplot as plt 
 from get_data import get_canton_data 
 from sqlalchemy import create_engine
-engine = create_engine("postgresql://epigraph:epigraph@localhost:5432/epigraphhub")
-
+import config
+engine = create_engine(config.DB_URI)
 
 def make_inference_sir():
     Prev_post = pd.read_sql_table('prev_post', engine, schema = 'switzerland', index_col = 'datum')

bayesian_inference.py → app/bayesian_inference.py

@@ -6,14 +6,15 @@
 @author: eduardoaraujo
 """
 #import arviz as az
+
 import plotly.graph_objects as go
 import pandas as pd 
 import streamlit as st 
 from PIL import Image
 from get_data import get_canton_data 
 from sqlalchemy import create_engine
-
-engine = create_engine("postgresql://epigraph:epigraph@localhost:5432/epigraphhub")
+import config
+engine = create_engine(config.DB_URI)
 
 
 

clustering.py → app/clustering.py

@@ -8,7 +8,7 @@
 
 import streamlit as st 
 from get_data import compute_clusters
-from geoanalysis import correlation_map
+from geoanalysis import correlation_map, correlation_map_bokeh
 
 
 def app():
@@ -44,5 +44,8 @@ def app():
     curve = st.selectbox("On which series you want to base the correlation on?",
                 ('cases', 'hospitalizations') 
     )
-    figmap = correlation_map(curve=curve)
-    st.pyplot(figmap)
+    #figmap = correlation_map(curve=curve)
+    #st.pyplot(figmap)
+
+    figmap_bokeh = correlation_map_bokeh(curve=curve)
+    st.bokeh_chart(figmap_bokeh)

app/config.py

@@ -0,0 +1,14 @@
+import os
+from dotenv import load_dotenv
+load_dotenv('../.env')
+
+POSTGRES_USER = os.getenv("POSTGRES_USER")
+POSTGRES_PASSWORD = os.getenv("POSTGRES_PASSWORD")
+POSTGRES_HOST = os.getenv("POSTGRES_HOST")
+POSTGRES_PORT = os.getenv("POSTGRES_PORT")
+POSTGRES_DB = os.getenv("POSTGRES_DB")
+
+DB_URI = (
+     f"postgresql://{POSTGRES_USER}:{POSTGRES_PASSWORD}"
+     f"@{POSTGRES_HOST}:{POSTGRES_PORT}/{POSTGRES_DB}"
+)

dashboard.py → app/dashboard.py

@@ -6,8 +6,6 @@
 """
 
 import streamlit as st
-import bayesian_inference
-import SIR_forecast
 import forecast
 import for_other_cantons 
 import credits

for_other_cantons.py → app/for_other_cantons.py

@@ -14,8 +14,8 @@
 from get_data import get_curve
 from sqlalchemy import create_engine
 from plots import scatter_plot_cases_hosp_all
-engine = create_engine(
-    "postgresql://epigraph:epigraph@localhost:5432/epigraphhub")
+import config
+engine = create_engine(config.DB_URI)
 
 dict_cantons_names = {
     'Uri (UR)': 'UR',

forecast.py → app/forecast.py

@@ -24,10 +24,8 @@
 from get_data import get_canton_data, get_curve
 from plots import scatter_plot_cases_hosp
 from sqlalchemy import create_engine
-engine = create_engine(
-    "postgresql://epigraph:epigraph@localhost:5432/epigraphhub")
-
-
+import config
+engine = create_engine(config.DB_URI)
 
 
 

app/geoanalysis.py

@@ -0,0 +1,97 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+from bokeh.models import (ColorBar,
+                          GeoJSONDataSource, HoverTool,
+                          LinearColorMapper)
+from bokeh.palettes import brewer
+from bokeh.plotting import figure
+
+from get_data import get_ch_map, lag_ccf, engine
+
+def correlation_map(curve='cases'):
+    if curve == 'hospitalizations':
+        curve = 'hosp'
+    chmap = get_ch_map()
+
+    df = pd.read_sql_table(f'foph_{curve}', engine, schema = 'switzerland', columns = ['datum','geoRegion',  'entries'])
+    df.index = pd.to_datetime(df.datum)
+    inc_canton = df.pivot(columns='geoRegion', values='entries')
+    del inc_canton['CHFL']
+    del inc_canton['CH']
+    cm, lm = lag_ccf(inc_canton.values)
+    corrsGE = pd.DataFrame(index= inc_canton.columns, data={'corr':cm[8,:]})
+    chmap_corr = pd.merge(left=chmap, right=corrsGE, on='geoRegion')
+    fig, ax = plt.subplots(1, 1)
+
+    chmap_corr.plot(column='corr', ax=ax,
+        legend=True,
+        legend_kwds={'label': "Correlation",
+                        'orientation': "horizontal"}
+    )
+    ax.set_axis_off();
+    return fig
+
+
+def correlation_map_bokeh(curve='cases'):
+    if curve == 'hospitalizations':
+        curve = 'hosp'
+    chmap = get_ch_map()
+
+    df = pd.read_sql_table(f'foph_{curve}', engine, schema = 'switzerland', columns = ['datum','geoRegion',  'entries'])
+    df.index = pd.to_datetime(df.datum)
+    inc_canton = df.pivot(columns='geoRegion', values='entries')
+    del inc_canton['CHFL']
+    del inc_canton['CH']
+    cm, lm = lag_ccf(inc_canton.values)
+    corrsGE = pd.DataFrame(index= inc_canton.columns, data={'corr':cm[8,:]})
+    chmap_corr = pd.merge(left=chmap, right=corrsGE, on='geoRegion')
+
+    # Input GeoJSON source that contains features for plotting
+    geosource = GeoJSONDataSource(geojson = chmap_corr.to_json())
+
+    # Define custom tick labels for color bar.
+    palette = brewer['BuGn'][8]
+    palette = palette[::-1] # reverse order of colors so higher values have darker colors
+    # Instantiate LinearColorMapper that linearly maps numbers in a range, into a sequence of colors.
+    color_mapper = LinearColorMapper(palette = palette, low = 0.6, high = 1.0)
+    # Define custom tick labels for color bar.
+    tick_labels = { '0.7': '0.7', '0.75': '0.75', '0.8': '0.8', '0.85': '0.85', '0.9': '0.9', '0.95': '0.95', '1.00': '1.00'}
+    # Create color bar.
+    color_bar = ColorBar(color_mapper = color_mapper, 
+                        label_standoff = 8,
+                        width = 500, height = 20,
+                        border_line_color = None,
+                        location = (0,0), 
+                        orientation = 'horizontal',
+                        major_label_overrides = tick_labels,
+                        title = 'Correlation')
+    # Create figure object.
+    p = figure(title = '', 
+            plot_height = 600, plot_width = 950, 
+            toolbar_location = 'below',
+            tools = "pan, wheel_zoom, box_zoom, reset")
+    p.xgrid.grid_line_color = None
+    p.ygrid.grid_line_color = None
+    # Add patch renderer to figure.
+    states = p.patches('xs','ys', source = geosource,
+                    fill_color = { 'field' :'corr',
+                                    'transform' : color_mapper},
+                    line_color = 'black', 
+                    line_width = 0.8, 
+                    fill_alpha = 1)
+    # Create hover tool
+    p.add_tools(HoverTool(renderers = [states],
+                        tooltips = [('Cantons','@geoRegion'),
+                                ('Correlation', '@corr')]))
+    # Specify layout
+    p.add_layout(color_bar, 'below')
+
+    p.xaxis.visible = False
+
+    p.yaxis.visible = False
+
+    return p
+
+
+
+

get_data.py → app/get_data.py

@@ -14,18 +14,16 @@
 import scipy.cluster.hierarchy as hcluster
 from sqlalchemy import create_engine
 import matplotlib.pyplot as plt
-#import streamlit as st
-
-engine_public = create_engine("postgresql://epigraph:epigraph@localhost:5432/epigraphhub")
-
+import config
+engine = create_engine(config.DB_URI)
 
 def get_curve_all(curve):
 
     '''
     Function to return the cases or hosp for all the switzerland
     '''
 
-    df2 = pd.read_sql(f"select datum, sum(entries) from switzerland.foph_{curve}_d group by datum;", engine_public)
+    df2 = pd.read_sql(f"select datum, sum(entries) from switzerland.foph_{curve}_d group by datum;", engine)
     df2.index = pd.to_datetime(df2.datum)
     df2 = df2.rename(columns={'sum': 'entries'})
     df2.sort_index(inplace = True)
@@ -45,7 +43,7 @@ def get_curve(curve, canton):
     dict_dates = {'cases': 'datum', 'hosp': 'datum', 
                   'hospcapacity': 'date'}
 
-    df = pd.read_sql(f"select {dict_columns[curve]} from switzerland.foph_{curve}_d where georegion='{canton}';", engine_public)
+    df = pd.read_sql(f"select {dict_columns[curve]} from switzerland.foph_{curve}_d where georegion='{canton}';", engine)
 
     df.index = pd.to_datetime(df[dict_dates[curve]])
 
@@ -104,7 +102,7 @@ def compute_clusters(curve, t, plot=False):
     -> all_regions: is the array with all the regions
     '''
 
-    df = pd.read_sql_table(f'foph_{curve}_d', engine_public, schema = 'switzerland', columns = ['datum','georegion',  'entries'])
+    df = pd.read_sql_table(f'foph_{curve}_d', engine, schema = 'switzerland', columns = ['datum','georegion',  'entries'])
 
     df.index = pd.to_datetime(df.datum)
 
@@ -171,7 +169,7 @@ def get_canton_data(curve, canton, ini_date=None):
                  }
 
     # getting the data from the databank
-    df = pd.read_sql_table(f'foph_{curve}_d', engine_public, schema = 'switzerland', columns = dict_cols[curve])
+    df = pd.read_sql_table(f'foph_{curve}_d', engine, schema = 'switzerland', columns = dict_cols[curve])
 
     # filtering by cantons
     df = df.loc[(df.georegion.isin(canton))]
@@ -190,6 +188,6 @@ def get_canton_data(curve, canton, ini_date=None):
 
 def get_ch_map():
     chmap = gpd.read_postgis("select * from switzerland.map_cantons;", 
-                    con=engine_public, geom_col="geometry")
+                    con=engine, geom_col="geometry")
     chmap['geoRegion'] = chmap.HASC_1.transform(lambda x: x.split('.')[-1])
     return chmap

plots.py → app/plots.py

@@ -65,6 +65,10 @@ def scatter_plot_cases_hosp(region):
     sns.regplot(x=cases['2022-01-01':'2022-03-31'].entries,
                     y=hosp['2022-01-01':'2022-03-31'].entries,
                     label='2022-Q1', color='tab:green',robust=True, ax=ax)
+
+    sns.regplot(x=cases['2022-04-01':'2022-06-30'].entries,
+                    y=hosp['2022-04-01':'2022-06-30'].entries,
+                    label='2022-Q2', color='tab:red',robust=True, ax=ax)
 
     ax.set_title('Cases vs hospitalizations in Geneva')
 

conda/dev.yaml

@@ -0,0 +1,14 @@
+name: env-covidch
+channels:
+  - conda-forge
+  - nodefaults
+dependencies:
+# - docker-compose
+  - pip
+  - gdal
+  - arviz
+  - geopandas
+  - psycopg2
+  - statsmodels
+  - openpyxl
+

docker/Dockerfile

@@ -0,0 +1,51 @@
+FROM condaforge/mambaforge
+
+USER root 
+
+ENV DEVIAN_FRONTEND=noninteractive
+
+ENV HOME "/home/covidch"
+ENV PATH "$PATH:/home/covidch/.local/bin"
+ENV ENV_NAME base
+ENV PATH "/opt/conda/envs/$ENV_NAME/bin:$PATH"
+ENV PATH "/opt/poetry/bin:$PATH"
+
+RUN apt-get -qq update --yes \
+  && apt-get -qq install --yes --no-install-recommends \
+  ca-certificates sudo curl \
+  && rm -rf /var/lib/apt/lists/*
+
+
+RUN useradd -ms /bin/bash covidch \
+  && echo "covidch ALL=(ALL) NOPASSWD: ALL" > /etc/sudoers.d/covidch \
+  && chmod 0440 /etc/sudoers.d/ \
+  && echo 'source /opt/conda/bin/activate "$ENV_NAME" && exec "$@"' > /activate.sh \
+  && echo 'source activate "$ENV_NAME"' >  /home/covidch/.bashrc \
+  && chmod +x /activate.sh \
+  && chmod -R a+rwx /opt/conda /tmp \
+  && sudo chown -R covidch:covidch /usr/src
+
+
+COPY --chown=covidch:covidch conda/dev.yaml /tmp/dev.yaml
+COPY --chown=covidch:covidch pyproject.toml poetry.lock README.md /usr/src/
+COPY --chown=covidch:covidch app /home/covidch/app/
+
+RUN curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/install-poetry.py | POETRY_HOME=/opt/poetry python && \
+    cd /usr/local/bin && \
+    ln -s /opt/poetry/bin/poetry && \
+    poetry config virtualenvs.create false
+
+RUN chmod -R a+rwx /home/covidch/.config/pypoetry/
+
+USER covidch
+
+RUN mamba env update -n $ENV_NAME \
+    --file /tmp/dev.yaml \
+  && cd /usr/src && poetry install \
+  && mamba clean -afy
+
+WORKDIR /home/covidch/app/
+
+ENTRYPOINT ["streamlit", "run", "dashboard.py"]
+
+#ENTRYPOINT ["bash", "/activate.sh"]

docker/docker-compose.yaml

@@ -0,0 +1,17 @@
+version: "3.9"
+services:
+  streamlit-app:
+    build:
+      context: ".."
+      dockerfile: docker/Dockerfile
+    env_file:
+      - ../.env
+    #environment:
+    #  - PSQL_HOST=${PSQL_HOST}
+    #  - PSQL_PORT=${PSQL_PORT}
+    ports:
+      - "8501:8501"
+    volumes:
+       # Mounting this file allows updates to the webpage without restarting the container
+      - ./:/app/
+