def get_data():

# Get treasury yield data from Quandl

df_treasury = quandl.get("USTREASURY/YIELD", authtoken=at.auth_key)

# Get Freddie, conforming and Jumbo 30 yr mortgage data from Quandl

# Get Freddie Mac Data

df_freddie_mac = quandl.get("FMAC/30US", authtoken=at.auth_key)

# Get Wells Fargo conforming data

df_wfc_conf_30 = quandl.get("WFC/PR_CON_30YFIXED_IR",

authtoken=at.auth_key)

# Get Wells Fargo Jumbo data

df_wfc_jumbo_30 = quandl.get("WFC/PR_JUMBO_30YFIXED_IR",

authtoken=at.auth_key)

# pickle the files

pickle.dump(df_treasury, open('treasury.pickle', 'wb'))

pickle.dump(df_freddie_mac, open('freddie_mac.pickle', 'wb'))

pickle.dump(df_wfc_conf_30, open('conf.pickle', 'wb'))

pickle.dump(df_wfc_jumbo_30, open('jumbo.pickle', 'wb'))

return df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30

def pickle_out():

df_treasury = pickle.load(open('treasury.pickle', 'rb'))

df_freddie_mac = pickle.load(open('freddie_mac.pickle', 'rb'))

df_wfc_conf_30 = pickle.load(open('conf.pickle', 'rb'))

df_wfc_jumbo_30 = pickle.load(open('jumbo.pickle', 'rb'))

return df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30

def manipulate_mortgage_data(df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30):

# Rename columns

df_freddie_mac.columns = ['Freddie Mac']

df_wfc_conf_30.columns = ['WFC Conf Int Rate']

df_wfc_jumbo_30.columns = ['WFC Jumbo Int Rate']

# Merge freddie mac, Wells Fargo conforming and Wells Fargo jumbo datafrmes

df_mortgage = pd.merge(df_freddie_mac, df_wfc_jumbo_30,

left_index = True, right_index = True)

df_mortgage = pd.merge(df_mortgage, df_wfc_conf_30,

left_index = True, right_index = True)

# Get differences in rates. Add them to the dataframe

df_mortgage['Conf minus Freddie'] = (df_mortgage['WFC Conf Int Rate'] - \

df_mortgage['Freddie Mac']).round(3)

df_mortgage['Jumbo minus Freddie'] = (df_mortgage['WFC Jumbo Int Rate'] - \

df_mortgage['Freddie Mac']).round(3)

df_mortgage.reset_index(inplace=True)

df_mortgage = df_mortgage.sort_values(by=['Date'], ascending=False)

return df_mortgage

def manipulate_yield_data(df_treasury):

# Take out 2 month yield and only use 24 months of data

df_treasury = df_treasury.drop(['2 MO'], axis=1).last('24M')

df_treasury.reset_index(inplace=True)

# Get the data for the last day of each business quarter for the

# last 24 months.

df_us_treasury = df_treasury.loc[df_treasury.Date.isin(df_treasury.Date + BQuarterEnd())]

# Once every 5 or 6 years the end of the first business

# quarter is Good Friday. The financial markets in the US

# are closed on Good Friday. When this occurs we have to get

# data from the day before Good Friday.

# Create a list of the years that are in the df_treasury dataframe

df_dates = pd.to_datetime(df_treasury['Date']).dt.year.unique()

for date in df_dates:

# Calculate Good Friday. It's two days before Easter Sunday

goodfriday = easter(date) + timedelta(days = -2)

# Calculate the end of the business quarter for the quarter that

# Good Friday is in.

Bqtr_end_date = (pd.to_datetime(goodfriday) + BQuarterEnd(0)).date()

# check to see if Good Friday is the last day of the business quarter

if goodfriday == Bqtr_end_date:

# Subtract one day from Good Friday to get financial end of qtr

end_of_qtr = pd.to_datetime(goodfriday + timedelta(days = -1))

# Get the row in df_treasury with the information that we need

df_temp = df_treasury[df_treasury.Date == end_of_qtr]

# Add the dataframe with the one record that we need to the

# dataframe with the business quarter end data

df_us_treasury = pd.concat([df_us_treasury, df_temp])

# Add data from most recent date in df_treasury to df_us_treasury

df_us_treasury = df_us_treasury.append(df_treasury.iloc[-1])

# Update the index so that the data prints out correctly

df_us_treasury = df_us_treasury.sort_values(by=['Date'], ascending=False)

df_treasury = df_treasury.sort_values(by=['Date'], ascending=False)

df_us_treasury['Date'] = pd.to_datetime(df_us_treasury['Date'], format = '%Y-%m-%d').dt.date

return df_us_treasury, df_treasury

def dash_create_charts(df_mortgage, df_us_treasury, df_treasury):

tabs_styles = {

'height': '44px',

'textAlign' : 'center'

}

tab_style = {

'borderBottom': '2px solid #d6d6d6',

'padding': '12px',

'fontWeight': 'bold'

}

tab_selected_style = {

'borderTop': '1px solid #d6d6d6',

'backgroundColor': '#119DFF',

'color': 'white',

'padding': '12px'

}

app.layout = html.Div([

html.H1(

'Interest Rate Information',

style = {'textAlign' : 'center'}

),

dcc.Tabs(

id= 'interest-rates',

value='mortgage-rates',

children=[

dcc.Tab(

label='Mortgage Rates',

value='mortgage-rates',

style=tab_style,

selected_style=tab_selected_style

),

dcc.Tab(

label='Mortgage Spreads',

value='mortgage-deltas',

style=tab_style,

selected_style=tab_selected_style

),

dcc.Tab(

label='Yield Curve',

value='yield-curve',

style=tab_style,

selected_style=tab_selected_style

),

], style = tabs_styles

),

html.Div(id='interest-rates-content')

])

@app.callback(Output('interest-rates-content', 'children'),

[Input('interest-rates', 'value')])

def render_content(tab):

if tab == 'mortgage-rates':

return (html.Div([

dash_table.DataTable(

id='datatable-int-rates',

data = df_mortgage.to_dict('records'),

columns = [

{

'id' : 'Date',

'name': 'Date',

'editable': False

},

{

'id' : 'Freddie Mac',

'name': 'Freddie Mac',

'editable': True

},

{

'id' : 'WFC Conf Int Rate',

'name': 'Wells Fargo Conforming Rates',

'editable': True

},

{

'id' : 'WFC Jumbo Int Rate',

'name': 'Wells Fargo Jumbo Rates',

'editable': True

},

],

virtualization = True,

fixed_rows = {

'headers' : True,

'data' : 0

},

style_cell = {

'textAlign' : 'center'

},

style_table = {

'maxHeight': '300px',

'overflowY' : 'scroll'

},

style_header = {

'backgroundColor': 'rgb(230,230,230)',

'fontWeight' : 'bold'

},

style_data_conditional=[

{'if': {'row_index': 'odd'}, 'backgroundColor': 'rgb(248, 248, 248)'},

],

),

html.Div(id = 'mortgage_graph')

]),

html.Div(

[

dcc.Graph(

id='fixed interest rates',

figure={

'data': [

{

'x': df_mortgage['Date'],

'y': df_mortgage['Freddie Mac'],

'type': 'line',

'name': 'Freddie Mac Interest Rates'

},

{

'x': df_mortgage['Date'],

'y': df_mortgage['WFC Conf Int Rate'],

'type': 'line',

'name': 'Wells Fargo Conforming Rates'

},

{

'x': df_mortgage['Date'],

'y': df_mortgage['WFC Jumbo Int Rate'],

'type': 'line',

'name': 'Wells Fargo Jumbo Rates'

},

],

'layout' : {

'title' : '<b>30 Year Fixed Rate Mortgages <br> Freddie Mac Vs. Wells Fargo<b>',

'xaxis' : {

'title' : '<b>Date<b>'

},

'yaxis' : {

'title' : '<b>Interest Rate<b>'

},

'height' : 500,

'margin' : {

"t": 80,

"l": 70,

"r": 20

},

'hovermode': 'closest',

},

}

)

]

)

)

elif tab == 'mortgage-deltas':

return (html.Div([

dash_table.DataTable(

id='datatable-deltas',

data = df_mortgage.to_dict('rows'),

columns = [

{

'id' : 'Date',

'name': 'Date',

'editable': False

},

{

'id' : 'Conf minus Freddie',

'name': 'Wells Fargo Conforming Rates minus Freddie',

'editable': True,

'type' : 'numeric',

'format' : Format(

precision = 3,

scheme = Scheme.fixed,

sign = Sign.parantheses),

},

{

'id' : 'Jumbo minus Freddie',

'name': 'Wells Fargo Jumbo Rates minus Freddie',

'editable': True,

'type' : 'numeric',

'format' : Format(

precision = 3,

scheme = Scheme.fixed,

sign = Sign.parantheses),

},

],

virtualization = True,

fixed_rows = {

'headers' : True,

'data' : 0

},

style_cell = {

'textAlign' : 'center'

},

style_header = {

'fontWeight' : 'bold',

'backgroundColor': 'rgb(230,230,230)',

},

style_table = {

'maxHeight': '300px',

'overflowY' : 'scroll'

},

style_data_conditional=[

{'if': {'row_index': 'odd'},

'backgroundColor': 'rgb(248, 248, 248)'},

],

),

html.Div(id = 'delta_graph')

]),

html.Div(

[

dcc.Graph(

id='fixed interest rates',

figure={

'data': [

{

'x': df_mortgage['Date'],

'y': df_mortgage['Conf minus Freddie'],

'type': 'line',

'name': 'Wells Fargo Conforming <br> Rates minus Freddie'

},

{

'x': df_mortgage['Date'],

'y': df_mortgage['Jumbo minus Freddie'],

'type': 'line',

'name': 'Wells Fargo Jumbo <br> Rates minus Freddie'

},

],

'layout' : {

'title' : '<b>Spreads in Mortgage Interest Rates<br>30 Year Fixed Rate Mortgage<b>',

'xaxis' : {'title' : '<b>Date<b>'},

'yaxis' : {'title' : '<b>Interest Rate<b>'},

'height' : 500,

'margin' : {

"t": 80,

"l": 70,

"r": 20

},

'hovermode': 'closest',

},

}

)

]

)

)

# elif tab == 'yield-curve':

else:

return (html.Div([

dash_table.DataTable(

id='fixed interest rates',

data = df_treasury.to_dict('rows'),

columns=[{"name": i, "id": i} for i in df_treasury.columns],

virtualization = True,

page_action = 'none',

editable = False,

fixed_rows = {

'headers' : True, 'data' : 0

},

style_cell = {

'textAlign' : 'center'

},

style_header = {

'backgroundColor': 'rgb(230,230,230)',

'fontWeight' : 'bold'

},

style_table = {

'maxHeight': '300px',

'overflowY' : 'scroll'

},

style_data_conditional=[

{

'if': {'row_index': 'odd'},

'backgroundColor': 'rgb(248, 248, 248)'

}

],

style_cell_conditional = [

{

'if': {'column_id' : key}, 'width' : value

} for key, value in at.col_dict.items()

],

),

html.Div(id = 'yield_curve_data')

]),

html.Div(

[

dcc.Graph(

id='yield-curve-graph',

figure={

'data': [

{'x': df_us_treasury.iloc[i][0],

'y': df_us_treasury.iloc[i],

'type': 'line',

'name' : df_us_treasury.iloc[i][0],

} for i in range(len(df_us_treasury))

],

'layout' : {

'title' : '<b>Yield Curve End of Quarter - Eight Quarters Running<b>',

'xaxis' : {

'title' : '<b>Tenor<b>',

'tickmode' : 'array',

'tickvals' : [i for i in range(1, len(df_us_treasury.columns))],

'ticktext' :

[i for i in df_us_treasury.columns.tolist() if i != 'Date'],

'range' : ['1', len(df_us_treasury.columns)-1]

},

'yaxis' : {

'title' : '<b>Interest Rate<b>'

},

'height' : 500,

'margin' : {

"t": 80,

"l": 70,

"r": 20

},

'hovermode': 'closest',

},

}

)

]

)

)

## Run the Code. ##

# Get the data from Quandl or from pickle file

# # does the pickle file exist

# if os.path.exists(at.last_pickle_file):

# print('the file exists')

# # if the file exists get the last date that the file was run

# pickle_date = time.strftime("%Y-%m-%d", time.localtime(os.stat(at.last_pickle_file)[8]))

# # if the pickle date is less than today

# if pickle_date < datetime.today().strftime('%Y-%m-%d'):

# print('pickle date is not today')

# # if the pickle was not run on a Friday

# if datetime.strptime(pickle_date, '%Y-%m-%d').weekday() != 3:

# print('pickle date is not equal to today and its day is not Friday')

# # get updated data

# df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30 = get_data()

# else:

# # pickle was run on a Friday and we are ok

# print('the pickle day of the week is Friday')

# df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30 = pickle_out()

# else:

# if the pickle file does not exist, get the data from quandl and create the pickle files.

df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30 = get_data()

# df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30 = get_data()

# df_treasury, df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30 = pickle_out()

# Manipulate data for use in dash

df_mortgage = manipulate_mortgage_data(df_freddie_mac, df_wfc_conf_30, df_wfc_jumbo_30)

df_mortgage.Date = pd.DatetimeIndex(df_mortgage.Date).strftime("%Y-%m-%d")

df_us_treasury, df_treasury = manipulate_yield_data(df_treasury)

df_us_treasury.Date = pd.DatetimeIndex(df_us_treasury.Date).strftime("%Y-%m-%d")

df_treasury.Date = pd.DatetimeIndex(df_treasury.Date).strftime("%Y-%m-%d")

# Create datatables and graphs

dash_create_charts(df_mortgage, df_us_treasury, df_treasury)

print('Complete main program')