reverse colorbar #65
reverse colorbar #65
Conversation
|
I can get the plot to work, but I think it's best to change up the code so that we put in a place holder for the date. I think the dashboard can go and handle the animation instead of plotly directly. See code comment for changes I made to make the iteration process faster |
chendaniely
left a comment
chendaniely
left a comment
There was a problem hiding this comment.
If you change you code to this, it should at least plot faster...
analysis/db/us_map/choroplethMap.py
Outdated
| # reversed_viridis = color_map.reversed() | ||
|
|
||
|
|
||
| fig = px.choropleth(molten_df, |
There was a problem hiding this comment.
I created a dataframe that was just a subset of a particular date, and then used that subseted dataframe to plot the figure
plot_data = molten_df[molten_df.date_iso == '2020-02-01']
fig = px.choropleth(plot_data,
geojson=counties,
locations=plot_data.fips_str,
color='value',
#animation_frame='date',
hover_data=['State', 'value'],
color_continuous_scale='viridis_r',
range_color=(0, 300),
scope="usa",
title='Confirmed cases',
labels={'value': 'confirmed cases'}
)
analysis/db/us_map/choroplethMap.py
Outdated
| @@ -34,7 +34,10 @@ | |||
| molten_df['date_iso'] = pd.to_datetime(molten_df['date'], format="%m/%d/%y") # change date to ISO8601 standard format | |||
|
|
|||
| fips = molten_df['fips_str'].tolist() | |||
There was a problem hiding this comment.
Because of the below changes, you don't need this line anymore since you're passing in the column of values into the plotting function
| confirmed_df = pd.read_csv('https://github.com/CSSEGISandData/COVID-19/raw/master/csse_covid_19_data/' | ||
| 'csse_covid_19_time_series/time_series_covid19_confirmed_US.csv') | ||
| loc_df = pd.read_excel(here('./data/db/original/maps/State_FIPS.xlsx')) | ||
| pop_df = pd.read_excel(here('./data/db/original/maps/PopulationEstimates.xls')) # population dataset for 2019 |
There was a problem hiding this comment.
where did this dataset come from?
| 'csse_covid_19_time_series/time_series_covid19_confirmed_US.csv') | ||
| loc_df = pd.read_excel(here('./data/db/original/maps/State_FIPS.xlsx')) | ||
| pop_df = pd.read_excel(here('./data/db/original/maps/PopulationEstimates.xls')) # population dataset for 2019 |
There was a problem hiding this comment.
should provide a download link to where you got these datasets from.
analysis/db/us_map/choroplethMap.py
Outdated
|
|
||
| molten_pop_df = pd.merge(molten_df, pop_df, on='fips_str') # add population per county | ||
| grouped_by = molten_pop_df.groupby(['fips_str', 'date_iso', 'Admin2', 'POP_ESTIMATE_2019'])['value'].sum().reset_index() | ||
| grouped_by['value'] = grouped_by['value']/grouped_by['POP_ESTIMATE_2019'] # get per capita value |
There was a problem hiding this comment.
don't overwrite the original 'value' column. you should make a new column (in this case something like 'total_per_cap') that is assigned the per capita value
analysis/db/us_map/choroplethMap.py
Outdated
| color_continuous_scale="Viridis", | ||
| range_color=(0, 300), | ||
| color_continuous_scale='viridis_r', | ||
| range_color=(0, 500), |
There was a problem hiding this comment.
why did you choose 500? can we set this to something like max(per_cap) and use a variable instead of hard-coding a value?
There was a problem hiding this comment.
Yea, you are right and I am working on that. I was thinking of putting there the third quartile as 75%, cause when I am taking the max value, which is for New York, it is much higher than other states and that's why it gets a bit wrong coloring. I tried to use quartile's fuction, but range_color didn't accept my input. The same goes with per capita case, but there it shows another state with the highest cases number, which is very strange, so I am assuming that I might be doing wrong calculations
| ''' | ||
| # ax = sns.lineplot(x="date_iso", y="value", hue='Province_State', data=grouped_counts) # show cases per state monthly | ||
| # ax = sns.stripplot(x="date_iso", y="value", hue='Province_State', data=grouped_counts) | ||
| # ax = sns.violinplot(x='date_iso', y='value', hue='Province_State', data=grouped_counts, palette="Set2", split=True, | ||
| # scale="count", inner="quartile") | ||
| # ax = sns.countplot(x="date_iso", hue='Province_State', data=grouped_counts) # works better if there are certain dates | ||
| # plt.tight_layout() | ||
| # plt.show() | ||
| ''' |
There was a problem hiding this comment.
why did you comment these out? we could also add general values into the dashboard too
analysis/db/us_map/choroplethMap.py
Outdated
| # animation_frame='date', | ||
| hover_data=['Admin2', 'value', 'POP_ESTIMATE_2019'], | ||
| color_continuous_scale='viridis_r', | ||
| range_color=(0, plot_data['value'].max()), |
There was a problem hiding this comment.
when you use the new column variable name make sure you change this as well.
analysis/db/us_map/choroplethMap.py
Outdated
|
|
||
|
|
||
|
|
||
| ''' No newline at end of file |
There was a problem hiding this comment.
Files should end with a new line
Also. might be worth having the raw count, and also the per-capita count as a toggle between the maps.
Since the only real difference between the plotting code is which column you're using to plot, we can make a function that takes a dataframe, and plotting column as input and returns the plot.
Would be able to use the function to return both plots that we would feed into the dashboard.
|
|
||
| confirmed_df = pd.read_csv('https://github.com/CSSEGISandData/COVID-19/raw/master/csse_covid_19_data/' | ||
| 'csse_covid_19_time_series/time_series_covid19_confirmed_US.csv') | ||
| loc_df = pd.read_excel(here('./data/db/original/maps/State_FIPS.xlsx')) |
There was a problem hiding this comment.
link to where you got data from
No description provided.