Olá Stack Exchange,
Para a vida de mim eu não posso imaginar o que eu estou fazendo de errado neste ponto. Eu sou um programador novato então, eu estou orgulhoso de chegar aqui; no entanto, se há outras soluções que são mais eficazes, por favor, sinta-se livre para sugerir.
Arquivos De Dados:
- gdf_final_serial
- ↑ Esta é uma versão resumida como o real conjunto de dados é ~680K Linhas
- geo_school
- ^Json sendo usado
O que eu preciso o Traço app para o fazer:
- Trabalho como Multi-Filtráveis Mapa SIG no que diz respeito a campos listados (município, Distrito, o Nome, Idade, RE, Quartis, Habitação Inseguro)
- De acordo com os filtros escolhidos, a atualização de px.choropleth_mapbox com a contagem precisa de 'pontos'.
Vamos para o código. Eu também estou executando isso no Google Colab Notebook.
filename_1 = root_path+"schoolDistrictBoundaries_Fixed.json"
file = open(filename_1)
schoolDistricts = gpd.read_file(file)
^^ sendo executados em uma célula global
import dash
import dash.dependencies
import plotly.express as px
import pandas as pd
from jupyter_dash import JupyterDash
import dash_core_components as dcc
import dash_html_components as html
from dash.dependencies import Input, Output
import plotly.graph_objects as go
import numpy as np
px.set_mapbox_access_token(mapbox_access_token)
### Creating gdf_final_serial
geo_df = gpd.GeoDataFrame.from_features(geo_school["features"]).merge(joined, on="OBJECTID").set_index("OBJECTID") #GeoJson Join to Joined Dataframe
cleaned_geo_df = geo_df[["geometry_x", "CountyName_x", "RE", "QUARTILES", "HOUSING_INSECURE", "County", "Age", "DistrictName_x"]] #Picks the right columns
geo_df_final = cleaned_geo_df.rename(columns={"CountyName_x": "CountyName", "geometry_x": "geometry", 'DistrictName_x': 'DistrictName'}) #Changing Column Names
geo_df_final.to_pickle(root_path+"geo_df_final") #Write to drive as a pickle to serialize
df_final_serial = pd.read_pickle(root_path+"geo_df_final") #Read as DataFrame
gdf_final_serial = gpd.GeoDataFrame(data = df_final_serial, geometry= "geometry", crs = "epsg:4326") #Turn into GeoPandas DataFrame and set the geometry and crs
re_indicators = gdf_final_serial.RE.unique()
county_indicators = gdf_final_serial.CountyName.unique()
age_indicators = gdf_final_serial.Age.unique()
district_indicators = gdf_final_serial.DistrictName.unique()
external_stylesheets = ["https://codepen.io/chriddyp/pen/bWLwgP.css"]
app = JupyterDash(__name__, external_stylesheets=external_stylesheets)
app.layout = html.Div(
[
html.Div(
children=[
html.Label("County"),
dcc.Checklist(
id="county",
options=[{"label": i, "value": i} for i in county_indicators],
value=[]
),
html.Label("District Name"),
dcc.Dropdown(
id="dname",
options=[],
value=[],
multi=True,
),
html.Label("Age"),
dcc.Dropdown(id="age",
options=[],
value=[],
multi = True),
html.Label("RE"),
dcc.Dropdown(id="re",
options=[],
value=[],
multi = True),
html.Label("Quartiles"),
dcc.Dropdown(id="quartiles",
options=[],
value=[],
multi=True),
html.Label("Housing"),
dcc.Dropdown(id="housing",
options=[],
value=[],
multi=True),
dcc.Graph(id="my_map", figure={})])
]
)
@app.callback(
dash.dependencies.Output("dname", "options"),
dash.dependencies.Input("county", "value")
)
def choose_county(county_pick):
if len(county_pick) > 0:
dff=gdf_final_serial[gdf_final_serial.CountyName.isin(county_pick)]
else:
raise dash.exceptions.PreventUpdate
return [{"label": i, "value": i} for i in (dff.DistrictName.unique())]
@app.callback(
dash.dependencies.Output("dname", "value"),
dash.dependencies.Input("dname", "options"),
)
def set_city_value(available_options_dname):
return [x["value"] for x in available_options_dname]
@app.callback(
dash.dependencies.Output("age", "options"),
dash.dependencies.Input("dname", "value")
)
def dname_age_picker(choose_dname):
print(choose_dname)
if len(choose_dname) > 0:
dff = gdf_final_serial[gdf_final_serial.DistrictName.isin(choose_dname)]
else:
raise dash.exceptions.PreventUpdate
return [{"label": i, "value": i} for i in (dff.Age.unique())]
@app.callback(
dash.dependencies.Output("age", "value"),
dash.dependencies.Input("age", "options"),
)
def set_age_value(available_options_age):
return [x["value"] for x in available_options_age]
@app.callback(
dash.dependencies.Output("re", "options"),
dash.dependencies.Input("age", "value")
)
def age_re_picker(choose_age):
if len(choose_age) > 0:
dff = gdf_final_serial[gdf_final_serial.Age.isin(choose_age)].dropna(axis = 0, how = 'any', subset = ["RE"])
else:
raise dash.exceptions.PreventUpdate
return [{"label": i, "value": i} for i in (dff.RE.unique())]
@app.callback(
dash.dependencies.Output("re", "value"),
dash.dependencies.Input("re", "options")
)
def set_re_value(available_options_re):
return [x["value"] for x in available_options_re]
@app.callback(
dash.dependencies.Output("quartiles", "options"),
dash.dependencies.Input("re", "value")
)
def re_quartile_picker(choose_re_value):
if len(choose_re_value) >= 0:
dff = gdf_final_serial[gdf_final_serial.RE.isin(choose_re_value)].dropna(axis = 0, how = 'any', subset = ["QUARTILES"])
else:
raise dash.exceptions.PreventUpdate
return [{"label": i, "value": i} for i in (dff.QUARTILES.unique())]
@app.callback(
dash.dependencies.Output("quartiles", "value"),
dash.dependencies.Input("quartiles", "options"),
)
def set_quart_value(available_options_quart):
return [x["value"] for x in available_options_quart]
@app.callback(
dash.dependencies.Output("housing", "options"),
dash.dependencies.Input("quartiles", "value")
)
def quart_picker(choose_quart_value):
if len(choose_quart_value) >= 0:
dff = gdf_final_serial[gdf_final_serial.QUARTILES.isin(choose_quart_value)]
else:
raise dash.exceptions.PreventUpdate
return [{"label": i, "value": i} for i in (dff.HOUSING_INSECURE.unique())]
@app.callback(
dash.dependencies.Output("housing", "value"),
dash.dependencies.Input("housing", "options"),
)
def set_housing_value(available_options_housing):
return [x["value"] for x in available_options_housing]
@app.callback(
dash.dependencies.Output("my_map", "figure"),
[dash.dependencies.Input("housing", "value"),
dash.dependencies.Input("quartiles", "value"),
dash.dependencies.Input("re", "value"),
dash.dependencies.Input("age", "value"),
dash.dependencies.Input("dname", "value"),
dash.dependencies.Input("county", "value")]
)
def update_fig(selected_housing, selected_quartiles, selected_re, selected_age, selected_dname, selected_county):
gdff_1 = gdf_final_serial[gdf_final_serial.CountyName.isin(selected_county) &
gdf_final_serial.DistrictName.isin(selected_dname) &
gdf_final_serial.Age.isin(selected_age) &
gdf_final_serial.RE.isin(selected_re) &
gdf_final_serial.QUARTILES.isin(selected_quartiles) &
gdf_final_serial.HOUSING_INSECURE.isin(selected_housing)]
count_points = gdff_1.groupby("OBJECTID").size().rename("points")
gdf_last = gdff_1.merge(count_points, on="OBJECTID", how="right", right_index = True)
gdf_last.to_pickle(root_path+"gdf_last") #Write to drive as a pickle to serialize
ddf_final_serial = pd.read_pickle(root_path+"gdf_last") #Read as DataFrame
gddf_final_serial = gpd.GeoDataFrame(data = ddf_final_serial, geometry= "geometry", crs = "epsg:4326")
gdff_2 = gddf_final_serial.head(50)
px.set_mapbox_access_token(mapbox_access_token)
fig = px.choropleth_mapbox(data_frame= gdff_2,
geojson= geo_school,
locations= 'DistrictName',
featureidkey = 'properties.DistrictName',
color="points",
center={"lat": 38.5941, "lon": -119.8815},
mapbox_style = 'dark').update_layout(mapbox_accesstoken = mapbox_access_token)
return fig
# Run app and display result inline in the notebook
if __name__ == "__main__":
app.run_server(host="127.0.0.1", port="8888",debug=True, use_reloader=False)
Eu tenho sido capaz de receber várias condições em que o mapa/filtro irá produzir um único concelho, às vezes dois municípios--, mas nunca a coisa toda. Tenha em mente o meu objetivo final é criar para o TODO o conjunto de dados (talvez usando RapidsAI da NVIDIA; no entanto, eu estou tendo problemas ao instalar o que bem no Google Colab).
O que é mais interessante é que, quando eu executar esse código fora do traço (como uma amostra do conjunto de dados reais) e passá-lo para a fig.show(), ele aparecerá; no entanto, ele não funcionará no Traço app.
Eu suspeito que o problema é com a minha json ou tratamento final de chamada de retorno como o depurador indica que, enquanto as entradas são de lá, my_map.a figura não é a saída de dados. Agradeço a ajuda.