In [ ]:
# uncomment to install in colab
# !pip install -e git+https://github.com/enzoampil/fastquant.git@master#egg=fastquant
In [1]:
symbol="JFC"
start_date="2018-01-01"
end_date="2020-04-01"
In [2]:
from fastquant import get_stock_data
data = get_stock_data(symbol=symbol,
start_date=start_date,
end_date=end_date,
source="phisix",
format="dc",
)
data.columns = "ds y".split()
data.head()
Out[2]:
| ds | y | |
|---|---|---|
| 1834 | 2018-01-03 | 255.4 |
| 1835 | 2018-01-04 | 255.0 |
| 1836 | 2018-01-05 | 255.0 |
| 1837 | 2018-01-08 | 256.0 |
| 1838 | 2018-01-09 | 255.8 |
In [3]:
from fastquant import DisclosuresPSE
psed = DisclosuresPSE(symbol=symbol,
start_date="01-01-2018",
end_date="01-04-2020",
)
Pulling JFC disclosures summary... 2 pages detected!
0%| | 0/59 [00:00<?, ?it/s]
Found 59 disclosures between 01-01-2018 & 01-04-2020 with 14 types: ['Press Release' 'Joint Ventures' 'Declaration of Cash Dividends' 'Change in Corporate Contact Details and/or Website' 'Material Information/Transactions' 'Substantial Acquisitions' 'Clarification of News Reports' "Results of Annual or Special Stockholders' Meeting" 'Results of Organizational Meeting of Board of Directors' "Notice of Annual or Special Stockholders' Meeting" 'Acquisition or Disposition of Shares of Another Corporation' 'Change in Directors and/or Officers (Resignation, Removal or Appointment, Election and/or Promotion)' 'Amendments to By-Laws' 'Amendments to Articles of Incorporation'] Pulling details in all JFC disclosures...
100%|██████████| 59/59 [01:25<00:00, 1.48s/it]
Loaded: /home/jp/github/research/project/fastquant/fastquant/../data/JFC_disclosures_1-1-2020_4-1-2020.csv Deleted: /home/jp/github/research/project/fastquant/fastquant/../data/JFC_disclosures_1-1-2020_4-1-2020.csv Saved: /home/jp/github/research/project/fastquant/fastquant/../data/JFC_disclosures_01-01-2018_01-04-2020.csv
In [4]:
psed
Out[4]:
DisclosuresPSE('JFC', 'all', '01-01-2018', '01-04-2020', True, True)
In [5]:
from fbprophet import Prophet
m = Prophet(growth="linear",
seasonality_mode='additive',
daily_seasonality=False,
weekly_seasonality=True,
yearly_seasonality=True,
interval_width=0.95, #uncertainty
holidays=None,
n_changepoints=20,
)
m.add_country_holidays(country_name='PH')
# m.add_seasonality(name='monthly', period=30.5, fourier_order=5)
m.fit(data)
Importing plotly failed. Interactive plots will not work.
Out[5]:
<fbprophet.forecaster.Prophet at 0x7fa64835d310>
In [6]:
m.train_holiday_names
Out[6]:
0 New Year's Day 1 Maundy Thursday 2 Good Friday 3 Day of Valor 4 Labor Day 5 Independence Day 6 Eid al-Fitr 7 Feast of the Sacrifice 8 National Heroes' Day 9 Bonifacio Day 10 Christmas Day 11 Rizal Day 12 Day of Valor, Maundy Thursday dtype: object
In [7]:
future = m.make_future_dataframe(periods=28)
future.tail()
Out[7]:
| ds | |
|---|---|
| 570 | 2020-04-25 |
| 571 | 2020-04-26 |
| 572 | 2020-04-27 |
| 573 | 2020-04-28 |
| 574 | 2020-04-29 |
In [8]:
forecast = m.predict(future)
forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail()
Out[8]:
| ds | yhat | yhat_lower | yhat_upper | |
|---|---|---|---|---|
| 570 | 2020-04-25 | 103.177183 | 83.587591 | 121.723567 |
| 571 | 2020-04-26 | 101.433078 | 82.308303 | 120.851018 |
| 572 | 2020-04-27 | 88.637509 | 70.148413 | 108.970685 |
| 573 | 2020-04-28 | 86.705508 | 67.810857 | 104.951139 |
| 574 | 2020-04-29 | 84.994734 | 66.316919 | 105.360896 |
In [9]:
fig1 = m.plot(forecast, uncertainty=True)
In [10]:
from fastquant import datestring_to_datetime
fig1 = m.plot(forecast)
#zoom
fig1.axes[0].set_xlim(datestring_to_datetime("2020-02-01"),
datestring_to_datetime("2020-04-30")
)
Out[10]:
(737456.0, 737545.0)
In [11]:
fig2 = m.plot_components(forecast)
In [12]:
from fbprophet.plot import add_changepoints_to_plot
fig = m.plot(forecast)
a = add_changepoints_to_plot(fig.gca(), m, forecast)
In [13]:
fig = psed.plot_disclosures(disclosure_type="substantial acquisitions")
cp = add_changepoints_to_plot(fig.gca(), m, forecast)
Pulling JFC stock data...
pm-prophet (under development)¶
In [ ]: