update

app.py

@@ -1,13 +1,17 @@
-import streamlit as st
-import pandas as pd
-import numpy as np
 import json
+import math
 import altair as alt
+import pandas as pd
+import streamlit as st
+st.set_page_config(layout="wide")
+
 from utils import get_variable_filter, is_numerical
 
+
 def check_codes(var, data_meta):
     return 'codes' in data_meta[var] and data_meta[var]['codes'] is not None and len(data_meta[var]['codes']) > 0
 
+
 @st.cache(show_spinner=False)
 def load_data():
     with st.spinner('loading data...'):
@@ -21,20 +25,24 @@ def load_data():
                 name2id[k] = {vv: int(kk) for kk, vv in v['codes'].items()}
         return data_main, data_meta, data_variable, data_filter, name2id
 
+
 DATA_MAIN, DATA_META, DATA_VARIABLE, DATA_FILTER, NAME2ID = load_data()
 
+
 def id2name(var, id_):
     id_ = str(id_)
     if var not in DATA_META or 'codes' not in DATA_META[var]:
         return id_
     return DATA_META[var]['codes'].get(id_, id_)
 
+
 @st.cache
 def name2id(var, name):
     if var in NAME2ID:
         return NAME2ID[var][name]
     return name
 
+
 def main():
     # st.write(df.head())
     load_data()
@@ -47,19 +55,22 @@ def main():
         'Heatmap': heat,
         'Histogram': hist,
     }
-    st.title('HRL Portal')
+    # st.title('HRL Portal')
     with st.sidebar:
-        chart = st.selectbox('Select a chart type:', list(charts.keys()))
+        st.subheader('Chart Type')
+        chart = st.selectbox('Select a chart type:', list(charts.keys()), label_visibility='collapsed')
     # st.write(DATA_VARIABLE)
     # st.write(DATA_FILTER)
     charts[chart]()
 
+
 @st.cache
 def get_unique(var):
     if check_codes(var, DATA_META):
         return list(DATA_META[var]['codes'].values())
     return DATA_MAIN[var].unique().tolist()
 
+
 def get_var_name(var):
     if var in DATA_META and 'name' in DATA_META[var]:
         name = DATA_META[var]['name']
@@ -68,75 +79,105 @@ def get_var_name(var):
         return f"{var} ({name})"
     return var
 
+
 def name2var(name):
     return name.split()[0]
 
+
 def meta():
     st.header('Variables')
     for k, v in DATA_META.items():
         st.subheader(k)
         st.write(v['description'])
 
+
 def area():
     with st.sidebar:
-        var_name = st.selectbox('Select a variable:', [get_var_name(var) for var in DATA_VARIABLE['area']])
+        st.subheader('Variable')
+        var_name = st.selectbox('Select a variable:', [get_var_name(var) for var in DATA_VARIABLE['area']], label_visibility='collapsed')
         var = name2var(var_name)
-        st.write('Filters:')
+        st.subheader('Filters')
+        # apply_fil = st.button('Apply Filters')
         filters = {}
         for fvar in DATA_FILTER[var]:
             if fvar == 'YEAR':
                 continue
             if is_numerical(fvar):
-                names = st.slider(fvar, min(get_unique(fvar)), max(get_unique(fvar)), value=(min(get_unique(fvar)), max(get_unique(fvar))))
+                names = st.slider(fvar, min(get_unique(fvar)), max(get_unique(fvar)),
+                                  value=(min(get_unique(fvar)), max(get_unique(fvar))))
             else:
                 names = st.multiselect(fvar, get_unique(fvar), default=get_unique(fvar))
             filters[fvar] = set([name2id(fvar, n) for n in names])
-    st.header(f'Area Chart: {var_name}')
-    with st.expander(f'{var}'):
-        st.write(DATA_META.get(var, {}).get('description', ''))
-
-    names = st.multiselect('Select values to display:', get_unique(var),
-                           default=['Authors',
-                                    'Musicians and music teachers',
-                                    'Telephone operators',
-                                    'Bus drivers','Cashiers'] if var == 'OCC1950' else ['Drugs and medicines',
-                                                                                        'Fisheries',
-                                                                                        'Glass and glass products',
-                                                                                        ])
-    vals = set([name2id(var, n) for n in names])
-    df = get_area_data(var, filters, vals)
+    st.header(f'Area Chart - {var_name}')
+    # with st.expander(f'{var}'):
+    # st.write(DATA_META.get(var, {}).get('description', ''))
+
+    # st.write(len([v for k, v in DATA_META[var]['codes'].items() if k != -1]))
+    # vals = set([name2id(var, n) for n in names])
+
+    container = st.container()
+    all = st.button("Select all")
+    clear = st.button("Clear all")
+
+    if all:
+        names = container.multiselect(f"Select one or more {var} values to display:",
+                                                 get_unique(var), [v for k, v in DATA_META[var]['codes'].items() if k != -1])
+    elif clear:
+        names = container.multiselect(f"Select one or more {var} values to display:",
+                                      get_unique(var), [])
+    else:
+        names = container.multiselect(f"Select one or more {var} values to display:",
+                                                 get_unique(var), default=['Authors', 'Musicians and music teachers', 'Telephone operators', 'Bus drivers', 'Cashiers'] if var == 'OCC1950' else ['Drugs and medicines','Fisheries','Glass and glass products',])
+
+    # if apply_fil:
+    #     df = get_area_data(var, filters)
+    # else:
+    #     df = get_area_data(var)
+    df = get_area_data(var, filters)
+    df = df[df[var].isin(names)]
     # st.write(df)
-    st.write(f'found {len(df)} records')
-    if len(df) > 0:
-        selection = alt.selection_multi(fields=[var], bind='legend')
-        plot = alt.Chart(df, title=f'Count of Different {var} Values').mark_area().encode(alt.X('YEAR'),
-                                                                                                         alt.Y('count', title='count', stack='zero'),
-                                                                                                         alt.Color(var,
-                                                                                                                   scale=alt.Scale(scheme='category20'),
-                                                                                                                   legend=alt.Legend(orient='bottom')),
-                                                                                          opacity=alt.condition(selection,
-                                                                                                                alt.value(
-                                                                                                                    1),
-                                                                                                                alt.value(
-                                                                                                                    0.3)),
-                                                                                          tooltip='count'
-                                                                                          )\
-            .properties(width=650).add_selection(selection)
-        st.write(plot)
+
+    if len(df) == 0:
+        st.warning('Select some data to display')
+        return
+    selection = alt.selection_multi(fields=[var], bind='legend')
+    plot = alt.Chart(df, title=f'Count of Different {var} Values').mark_area().encode(alt.X('YEAR'),
+                                                                                      alt.Y('count', title='count',
+                                                                                            stack='zero'),
+                                                                                      alt.Color(var,
+                                                                                                scale=alt.Scale(
+                                                                                                    scheme='category20'),
+                                                                                                legend=alt.Legend(
+                                                                                                    # orient='bottom',
+                                                                                                    # columns=3
+                                                                                                )),
+                                                                                      opacity=alt.condition(selection,
+                                                                                                            alt.value(
+                                                                                                                1),
+                                                                                                            alt.value(
+                                                                                                                0.3)),
+                                                                                      tooltip=['YEAR', var, 'count']
+                                                                                      ) \
+        .properties(height=400 + len(names)).add_selection(selection)
+    st.altair_chart(plot, use_container_width=True)
+
 
 @st.cache(show_spinner=False)
-def get_area_data(var, filters, vals):
+def get_area_data(var, filters=None):
     # for k, v in filters.items():
     #     st.write(k)
     #     st.write(v)
     #     st.write(DATA_MAIN[k].unique())
     # st.write(len(DATA_MAIN))
-    df = DATA_MAIN[['YEAR', var, *filters.keys()]]
-    df = df[df[var].isin(vals)]
-    with st.spinner('filtering...'):
-        for fvar, fvals in filters.items():
-            df = df[df[fvar].isin(fvals)]
-    df = df[['YEAR', var]]
+    if filters is not None:
+        df = DATA_MAIN[['YEAR', var, *filters.keys()]]
+        # df = df[df[var].isin(vals)]
+        with st.spinner('filtering...'):
+            for fvar, fvals in filters.items():
+                df = df[df[fvar].isin(fvals)]
+        df = df[['YEAR', var]]
+    else:
+        df = DATA_MAIN[['YEAR', var]]
     # st.write(len(df))
     # st.write(df.head())
     with st.spinner('counting...'):
@@ -144,24 +185,30 @@ def get_area_data(var, filters, vals):
         groups[var] = groups[var].apply(lambda x: id2name(var, x))
     return groups
 
+
 def line():
     st.header('Line Graph')
 
+
 def scatter():
     st.header('Scatter Plot')
 
+
 def bar():
     st.header('Bar Chart')
 
+
 def box():
     st.header('Box Plot')
 
+
 def heat():
     st.header('Heatmap')
 
+
 def hist():
     st.header('Histogram')
 
+
 if __name__ == '__main__':
     main()
-

data/meta.json

@@ -647,7 +647,7 @@
     },
     "OCC1950": {
         "name": "Occupation",
-        "description": "OCC1950 applies the 1950 Census Bureau occupational classification system to occupational data, to enhance comparability across years. For pre-1940 samples created at Minnesota, the alphabetic responses supplied by enumerators were directly coded into the 1950 classification. For other samples, the information in the variable OCC was recoded into the 1950 classification. Codes above 970 are non-occupational responses retained in the historical census samples or blank/unknown. The design of OCC1950 is described at length in \" Integrated Occupation and Industry Codes and Occupational Standing Variables in the IPUMS.\". The composition of the 1950 occupation categories is described in detail in U.S. Bureau of the Census, Alphabetic Index of Occupations and Industries: 1950 (Washington D.C., 1950).\n\nIn 1850-1880, any laborer with no specified industry in a household with a farmer is recoded into farm labor. In 1860-1900, any woman with an occupational response of \"housekeeper\" enters the non-occupational category \"keeping house\" if she is related to the head of household. Cases affected by these imputation procedures are identified by an appropriate data quality flag.\n\nA parallel variable called OCC1990, available for the samples from 1950 onward, codes occupations into a simplified version of the 1990 occupational coding scheme.",
+        "description": "OCC1950 applies the 1950 Census Bureau occupational classification system to occupational data, to enhance comparability across years. For pre-1940 samples created at Minnesota, the alphabetic responses supplied by enumerators were directly coded into the 1950 classification. For other samples, the information in the variable OCC was recoded into the 1950 classification.",
         "codes": {
             "0": "Accountants and auditors",
             "1": "Actors and actresses",