lazyplotly

A wrapper of interactive visualization package plotly. Sometimes we are just too lazy to make nested dictionary. Especially creating dropdown menus or slide bars is a way too exhuasting. Moreover, variables of different kind of chart are renamed into a single taxonomy.

What is simplified as easy as possible

1. create interactive widgets such as dropdown menu or a slider bar
2. combine different type of charts together into subplots
3. automatically calculate grid rows and columns of subplots
4. rename different type of variables to make them in a single taxonomy
5. initiate essential settings for jupyter notebook users

Installation

$ pip install plotly lazyplotly

Quick Start

import lazyplotly as lp

# using list data type to fit
bar_1 = lp.bar(x=[1,2,3,4,5], y=[23,43,62,24,33], cmap=lp.cmap[16], name='male')

# using dataframe to fit
df = dict(x=[1,2,3,4,5],y=[48,32,43,54,62])
bar_2 = lp.bar(x='x', y='y', df=df, cmap=lp.cmap[16], name='female')

# combine all charts into dropdown menu, and save as a html file
lp.dropdown(
    datas = [[bar_1,bar_2],[bar_2, bar_1]],
    btn_labels = ['male v.s. female','female v.s. male'], 
    xlabel = 'category', ylabel='value', title='MyAwesomeTitle',
    output = 'MyAwesomePlot.html'
)

Custom extension

# use dataframe object by setting xy variables with column name
df = dict(order=[1,2,3,4,5],trend=[22,33,47,57,46],lower=[21,32,43,54,32],upper=[23,34,51,60,60])
area = lp.area(x='order', y='lower', y2='upper', df=df, color='rgba(0,176,246,0.2)', name='CI')
line = lp.line(x='order', y='trend', df=df)
line['line'] = dict(dash = 'dash', color='steelblue')

# add output variable to export as a html file
lp.plot(
    data = [area, line],
    rangeslider = True,
    layout = dict(title='Time Series with confidence interval'), # layout extension
    config = dict(displayModeBar=True), # util icons on right top side
)

Subplot

import numpy as np
import pandas as pd
df = pd.DataFrame(dict(
    monday = np.random.normal(5, 1, 100),
    tuesday = np.random.normal(5, 1.5, 100),
    wednesday = np.random.normal(5, 2, 100),
    thursday = np.random.normal(5, 2.5, 100),
    friday = np.random.normal(5, 3, 100),
    saturday = np.random.normal(5, 3.5, 100),
    sunday = np.random.normal(5, 4, 100),
))
data = []
for x in df.columns:
    for y in df.columns:
        if x==y:
            data.append(lp.histogram(x=df[x], name=x))
        else:
            data.append(lp.scatter(x=df[x], y=df[y], name=f'{x}-{y}'))
# combine all subplots
lp.plot(data, rows=7)

# or make a collection of boxplot
lp.plot([lp.box(y=df[col], name=col) for col in df.columns])

Sunburst, Pie, Donut

df = dict(
    parents = ['', 'Eve', 'Eve', 'Seth', 'Seth', 'Eve', 'Eve', 'Awan', 'Eve' ],
    labels = ['Eve', 'Cain', 'Seth', 'Enos', 'Noam', 'Abel', 'Awan', 'Enoch', 'Azura'],
    values = [10, 14, 12, 10, 2, 6, 6, 4, 4]
)
sunburst = lp.pie(x='parents',x2='labels',y='values', df=df)
pie = lp.pie(x='labels', y='values', df=df)
donut = lp.pie(x='labels', y='values', df=df, hole=0.4)
lp.plot([sunburst, pie, donut], cols=3)

3D scatter

import numpy as np
df = dict(
    x=np.random.randint(100,size=100),
    y=np.random.randint(100,size=100),
    z=np.random.randint(100,size=100)
)
scatter3d = lp.scatter3d(x='x',y='y',z='z', df=df, cmap=lp.cmap[16])
mesh3d = lp.mesh3d(x='x',y='y',z='z', df=df, color=lp.color[0])
lp.plot([scatter3d,mesh3d], no_padding=True)

Sankey Diagram

# sankey diagram
df = dict(
    origin=['Eve', 'Cain', 'Abel', 'Abel', 'Noam'],
    destin=['Abel', 'Abel', 'Awan', 'Enoch', 'Abel'],
    count=[23,27,50,60,60]
)
sankey = lp.sankey(x='origin', y='destin', z='count', df=df)
lp.plot(sankey)

Map

token = 'your mapbox api access token, get yours at https://account.mapbox.com/'
df = dict(
    lon=[121.1,121.2,121.5,121.3],
    lat=[24.1,24.3,24.5,24.2],
    volume=[20,30,40,50],
    stop=['A','B','C','detail information about bus stop D']
)
scattermapbox = lp.scattermapbox(x='lon',y='lat',z='volume',tooltip='stop',name='bus route',df=df)
scattermapbox['mode'] = 'markers+lines'
lp.mapbox(scattermapbox,token=token,zoom=9)

APIs

lp.cmap     # show all colorscales available in plotly
lp.color    # show all defined css color name in plotly
plot(data=[], layout=dict, output=bool, config=dict,cols=int, rows=int,
    rangeslider=bool,no_padding=bool, title=str, xlabel=str, ylabel=str)
dropdown(datas=[[]], btn_labels=[], layout=dict, output=bool, config=dict,
    no_padding=bool, title=str, xlabel=str, ylabel=str)
slider(datas=[[]], prefix=str, layout=dict, output=bool, config=dict,
    no_padding=bool, title=str, xlabel=str, ylabel=str)
bar(x, y, df, name, color, cmap)
scatter(x, y, df, name, color)
scatter3d(x, y, z, df, name, color, cmap)
line(x, y, df, name, color)
line3d(x, y, z, df, name, color)
area(x, y, y2, df, name, color)
area3d(x, y, z, df, name, color)
mesh3d(x, y, z, df, name, color)
box(y, df, name, color)
histogram(x, df, name, color, <int>bins, <bool>cum, <bool>prob)
pie(x, x2, y, df, name, hole, color)
heatmap(df, cmap)
sankey(x, y, z, df)
scattermapbox(x, y, z, tooltip, df, name)
mapbox(data, token, zoom, output, config)