In [ ]:
#import library yang dibutuhkan

import pandas as pd
import numpy as np
from bokeh.io import output_notebook
from bokeh.plotting import figure, show, reset_output
from bokeh.io import output_file
from bokeh.models.widgets import Tabs, Panel
from bokeh.models import ColumnDataSource, CategoricalColorMapper, NumeralTickFormatter, HoverTool, CDSView, GroupFilter, Div
from bokeh.layouts import gridplot, column
In [ ]:
# mount to google drive
from google.colab import drive
drive.mount("/content/drive")

Mounted at /content/drive
In [ ]:
cd drive/MyDrive/Colab\ Notebooks/visdat/tubes interactive

/content/drive/MyDrive/Colab Notebooks/visdat/tubes interactive
In [ ]:
!ls

Andrew-1301180390.ipynb  stock_market.csv
In [ ]:
# membaca file 
DT = 'stock_market.csv'
dataset = pd.read_csv(DT, parse_dates=['Date'])
In [ ]:
# data exploration
dataset = dataset.rename(columns = {'Adj Close': 'AdjClose'}, inplace = False)
dataset.sort_values(by ='Date')
Out[ ]:
Date AdjClose Volume Day_Perc_Change Name
0 2018-10-22 26153.150390 2222812200 2.315014 HANG SENG
992 2018-10-22 22614.820310 63000 0.367211 NIKKEI
491 2018-10-22 7468.629883 2282400000 0.263123 NASDAQ
1 2018-10-23 25346.550780 1772912000 -3.084139 HANG SENG
993 2018-10-23 22010.779300 80200 -2.670996 NIKKEI
... ... ... ... ... ...
990 2020-10-15 11713.870120 3304490000 -0.466154 NASDAQ
991 2020-10-16 11671.559570 3156270000 -0.361200 NASDAQ
1472 2020-10-16 23410.630860 48100 -0.410936 NIKKEI
490 2020-10-16 24386.789060 2697525100 0.944801 HANG SENG
1473 2020-10-19 23671.130860 0 1.112742 NIKKEI

1474 rows × 5 columns

In [ ]:
# jumlah data
dataset.shape
Out[ ]:
(1474, 5)
In [ ]:
dataset.dtypes
Out[ ]:
Date               datetime64[ns]
AdjClose                  float64
Volume                      int64
Day_Perc_Change           float64
Name                       object
dtype: object
In [ ]:
# mengecek duplikasi data
dupeCheck = dataset.duplicated()
print('Duplikat data = %d' % (dupeCheck.sum()))

Duplikat data = 0
In [ ]:
# mengecek missing value
missingdata=dataset.isnull().sum().sort_values(ascending = False)
print(missingdata)

Name               0
Day_Perc_Change    0
Volume             0
AdjClose           0
Date               0
dtype: int64

membuat figure untuk adj close, volume, dan day percentage change, serta memilih tools yang akan tersedia nantinya

In [ ]:
# memnbuat figure untuk adjclose, volume, dan day percentage change
output_notebook() # menampilkan output pada notebook

# memilih tools
select_tools = ['pan', 'box_select', 'wheel_zoom', 'tap', 'reset']

FAdj = figure(x_axis_type='datetime', plot_height=500, plot_width=600, title='Adj Close', toolbar_location="below", tools=select_tools, x_axis_label='Date', y_axis_label='Adj Close')

FVolume = figure(x_axis_type='datetime', plot_height=500, plot_width=600, title='Volume', toolbar_location="below", tools=select_tools, x_axis_label='Date', y_axis_label='Volume')

FDayPerc = figure(x_axis_type='datetime', plot_height=500, plot_width=600, title='Day_Perc_Change', toolbar_location="below", tools=select_tools, x_axis_label='Date', y_axis_label='Day_Perc_Change')
In [ ]:
# mengubah dataset ke columdatasource
datasetFF = ColumnDataSource(dataset)
In [ ]:
# membuat view dan filter untuk masing-masing indeks saham
# hangseng
FILhangseng = [GroupFilter(column_name='Name', group='HANG SENG')]
VIEWhangseng = CDSView(source=datasetFF, filters=FILhangseng)
# nikkei
FILnikkei = [GroupFilter(column_name='Name', group='NIKKEI')]
VIEWnikkei = CDSView(source=datasetFF, filters=FILnikkei)
# nasdaq
FILnasdaq = [GroupFilter(column_name='Name', group='NASDAQ')]
VIEWnasdaq = CDSView(source=datasetFF, filters=FILnasdaq)
In [ ]:
# membuat dictionary untuk circle dan indeks saham
DICTcircle = {'source': datasetFF, 'size': 4, 'alpha': 0.7, 'selection_color':'red', 'nonselection_color':'lightgray', 'nonselection_alpha': 0.3, 'muted_alpha': 0.1} #circle
DICThangseng = {'view': VIEWhangseng, 'color': '#eb6777', 'legend': 'HANG SENG'} #hangseng
DICTnikkei = {'view': VIEWnikkei, 'color': '#e0e38f', 'legend': 'NIKKEI'} #nikkei
DICTnasdaq = {'view': VIEWnasdaq, 'color': '#073278', 'legend': 'NASDAQ'} #nasdaq
In [ ]:
# menambah data indeks saham dengan bentuk circle untuk figure
# utuk AdjClose
FAdj.circle(x='Date', y='AdjClose', **DICTcircle, **DICThangseng)
FAdj.circle(x='Date', y='AdjClose', **DICTcircle, **DICTnikkei)
FAdj.circle(x='Date', y='AdjClose', **DICTcircle, **DICTnasdaq)
# untuk Volume
FVolume.circle(x='Date', y='Volume', **DICTcircle, **DICThangseng)
FVolume.circle(x='Date', y='Volume', **DICTcircle, **DICTnikkei)
FVolume.circle(x='Date', y='Volume', **DICTcircle, **DICTnasdaq)
# untuk Day_Perc_Change
FDayPerc.circle(x='Date', y='Day_Perc_Change', **DICTcircle, **DICThangseng)
FDayPerc.circle(x='Date', y='Day_Perc_Change', **DICTcircle, **DICTnikkei)
FDayPerc.circle(x='Date', y='Day_Perc_Change', **DICTcircle, **DICTnasdaq)

BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
BokehDeprecationWarning: 'legend' keyword is deprecated, use explicit 'legend_label', 'legend_field', or 'legend_group' keywords instead
Out[ ]:
GlyphRenderer(
id = '1342', …)
data_source = ColumnDataSource(id='1095', ...),
glyph = Circle(id='1338', ...),
hover_glyph = None,
js_event_callbacks = {},
js_property_callbacks = {},
level = 'glyph',
muted = False,
muted_glyph = Circle(id='1341', ...),
name = None,
nonselection_glyph = Circle(id='1339', ...),
selection_glyph = Circle(id='1340', ...),
subscribed_events = [],
syncable = True,
tags = [],
view = CDSView(id='1101', ...),
visible = True,
x_range_name = 'default',
y_range_name = 'default')
In [ ]:
# untuk tooltips
adjTool = [('Name','@Name'), ('Adj Close', '@AdjClose')]
volumeTool = [('Name','@Name'), ('Volume', '@Volume')]
daypercTool = [('Name','@Name'), ('Day_Perc_Change', '@Day_Perc_Change')]

# untuk hover
adjHG = FAdj.circle(x='Date', y='AdjClose', source=datasetFF, size=7, alpha=0, hover_fill_color='red', hover_alpha=0.5)
volumeHG = FVolume.circle(x='Date', y='Volume', source=datasetFF, size=7, alpha=0, hover_fill_color='red', hover_alpha=0.5)
daypercHG = FDayPerc.circle(x='Date', y='Day_Perc_Change', source=datasetFF, size=7, alpha=0, hover_fill_color='red', hover_alpha=0.5)

# untuk hover glyph
FAdj.add_tools(HoverTool(tooltips=adjTool, renderers=[adjHG]))
FVolume.add_tools(HoverTool(tooltips=volumeTool, renderers=[volumeHG]))
FDayPerc.add_tools(HoverTool(tooltips=daypercTool, renderers=[daypercHG]))
In [ ]:
html = """<h3>Visualisasi Data</h3>
<br>
"""
judulUp = Div(text=html)

# agar dapat menyembunyikan data untuk indeks pasar saham tertentu
# dengan mengklik legend dari indeks tersebut
FAdj.legend.click_policy = 'hide'
FVolume.legend.click_policy = 'hide'
FDayPerc.legend.click_policy = 'hide'

# membuat panel yang untuk adj close
PAdjClose = Panel(child=FAdj, title='Adj Close')
# membuat panel yang untuk volume
PVolume = Panel(child=FVolume, title='Volume')
# membuat panel yang untuk adj day_perc_change
PDayPerc = Panel(child=FDayPerc, title='Day_Perc_Change')

# membuat panel-panel tersebut menjadi tab, nanti bisa diclik tiap tab
tabs = Tabs(tabs=[PAdjClose, PVolume, PDayPerc])

# menampilkan tab
show(column(judulUp,tabs))