import numpy as np
import plotly.graph_objects as go
from plotly.subplots import make_subplots

np.random.seed(1)
data0 = np.random.normal(10, 1, 10000)
data1 = np.random.normal(12, 1.5, 10000)

histogram2d_fig = make_subplots(rows=1, cols=2)
# 2次元ヒストグラム（左図）
histogram2d_fig.add_trace(go.Histogram2d(x=data0, y=data1), row=1, col=1)
# 等高線（右図）
histogram2d_fig.add_trace(
    go.Histogram2dContour(x=data0, y=data1, showscale=False), row=1, col=2
)
histogram2d_fig.update_layout(coloraxis_showscale=False)
histogram2d_fig.show()

probability_comulative_histogram_fig = make_subplots(rows=1, cols=2)
probability_comulative_histogram_fig.add_trace(
    go.Histogram(
        x=data0,
        histnorm="probability",	# サンプル数の合計を1とした正規化
        name="probability"
    ),
    row=1,
    col=1
)
probability_comulative_histogram_fig.add_trace(
    go.Histogram(
        x=data0,
        cumulative={"enabled": True},	# 累積ヒストグラム
        name="comulative"
    ),
    row=1,
    col=2
)
probability_comulative_histogram_fig.show()

go.Figure(
    go.Histogram(
        x=data0,
        # ヒストグラムの範囲を指定
        xbins={"start": 6, "end": 13, "size": 0.01}
    )
).show()

import numpy as np
import plotly.graph_objects as go
from plotly.subplots import make_subplots

np.random.seed(1)
data0 = np.random.normal(10, 1, 10000)
data1 = np.random.normal(12, 1.5, 10000)
histogram_fig = make_subplots(rows=1, cols=2)

# 左図
histogram_fig.add_trace(go.Histogram(x=data0, name="data0"), row=1, col=1)
histogram_fig.add_trace(go.Histogram(x=data1, name="data1"), row=1, col=1)

# 右図
histogram_fig.add_trace(go.Histogram(y=data0, name="data0"), row=1, col=2)
histogram_fig.add_trace(
    go.Histogram(
        y=data1,  	# 引数yで横向き
        name="data1",
        nbinsy=50   # ビンの数を変更
    ),
    row=1,
    col=2
)
histogram_fig.show()

go.Figure(
    [go.Histogram(x=data0, name="data0"), go.Histogram(x=data1, name="data1")],
    layout=go.Layout(barmode="stack"),  # ヒストグラムを積み上げ
).show()

import plotly
import plotly.graph_objects as go

tips = plotly.data.tips()

# 曜日ごとのデータを抽出
tips_by_day = tips.groupby("day")  # day列でグループ化
days = ["Thur", "Fri", "Sat", "Sun"]
# グループ化された各DataFrameを抽出
thur, fri, sat, sun = [tips_by_day.get_group(day) for day in days]
thur.head()

from plotly.subplots import make_subplots

violin_fig = make_subplots(rows=1, cols=2)
# 1列目のサブプロット
violin_fig.add_trace(
    go.Violin(
        y=thur["tip"],
        name="Thur",
        box_visible=True            # 箱ひげ図を重ねて描画
    ),
    row=1,
    col=1
)
violin_fig.add_trace(
    go.Violin(y=fri["tip"], name="Fri", box_visible=True), row=1, col=1
)
violin_fig.add_trace(
    go.Violin(y=sat["tip"], name="Sat", box_visible=True), row=1, col=1
)
violin_fig.add_trace(
    go.Violin(y=sun["tip"], name="Sun", box_visible=True), row=1, col=1
)

# 2列目のサブプロット
for data in thur, fri, sat, sun:
    smoker = data.loc[data["smoker"] == "Yes"]      # smoker列がYesのデータ
    non_smoker = data.loc[data["smoker"] == "No"]   # smoker列がNoのデータ
    day = data["day"].iloc[0]                       # day列の先頭のデータ(ラベルに使用)
    violin_fig.add_trace(
        go.Violin(
            x=smoker["day"],
            y=smoker["tip"],
            side="negative",        # 左側に描画
            name=f"{day}: somoker",
        ),
        row=1,
        col=2
    )
    violin_fig.add_trace(
        go.Violin(
            x=non_smoker["day"],
            y=non_smoker["tip"],
            side="positive",        # 右側に描画
            name=f"{day}: non-somoker",
        ),
        row=1,
        col=2
    )
violin_fig.show()

import plotly
import plotly.graph_objects as go

tips = plotly.data.tips()
tips.head()

# 曜日ごとのデータを抽出
tips_by_day = tips.groupby("day")  # day列でグループ化
days = ["Thur", "Fri", "Sat", "Sun"]

# グループ化された各DataFrameを抽出
thur, fri, sat, sun = [tips_by_day.get_group(day) for day in days]

box_fig = go.Figure()
box_fig.add_trace(go.Box(y=thur["tip"], name="Thur"))
box_fig.add_trace(go.Box(y=fri["tip"], name="Fri"))
box_fig.add_trace(go.Box(y=sat["tip"], name="Sat"))
box_fig.add_trace(go.Box(y=sun["tip"], name="Sun"))
box_fig.show()

import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots

table_values = [[10, 20, 30, 40], [34, 15, 27, 8]]
table_labels = ["Label1", "Label2"]
table_fig = make_subplots(
    rows=2,
    cols=2,
    specs=[[{"type": "domain"}, {"type": "domain"}], [{"colspan": 2}, None]],
)
table_fig.add_trace(
    go.Table(header={"values": table_labels}, cells={"values": table_values}),
    row=1,
    col=1,
)
table_fig.add_trace(
    go.Table(
        cells={
            "values": pd.DataFrame(table_values),
            "line": {"width": 2, "color": "black"},  # 罫線のスタイル
            "fill": {"color": "white"},  # 塗りつぶし色
            "align": "right",  # 配置
        },
        header={
            "values": table_labels,
            "height": 18,  # セルの高さ
            "line": {"width": 2, "color": "black"},
            "fill": {"color": "white"},
            "font": {"size": 10},  # フォントサイズ
        },
    ),
    row=1,
    col=2,
)
table_fig.add_trace(
    go.Scatter(x=table_values[0], y=table_values[1]), row=2, col=1
)
table_fig.show()

import plotly.graph_objects as go
from plotly.subplots import make_subplots

asset_labels  = ["資産", "債権", "A社",  "B社",  "株式", "C社",  "D社",  "預金","Z社"]
asset_parents = ["",     "資産", "債権", "債権", "資産", "株式", "株式", "資産","B社"]
asset_values  = [1000,    400,    300,    100,    200,    160,    40,     400,   50]

go.Figure(
    go.Treemap(
        labels=asset_labels,    # セクタごとのラベル
        parents=asset_parents,  # 親セクタのラベル
        values=asset_values,    # セクタごとの値
        branchvalues="total"    # 親が子の階層すべての合計値
    )
).show()

import plotly.graph_objects as go
from plotly.subplots import make_subplots

asset_labels  = ["資産", "債権", "A社",  "B社",  "株式", "C社",  "D社",  "預金","Z社"]
asset_parents = ["",     "資産", "債権", "債権", "資産", "株式", "株式", "資産","B社"]
asset_values  = [1000,    400,    300,    100,    200,    160,    40,     400,   50]

sunburst_fig = make_subplots(
    1, 2, specs=[[{"type": "domain"}, {"type": "domain"}]]
)

sunburst_fig.add_trace(
    go.Sunburst(
        labels=asset_labels,        # セクタごとのラベル
        parents=asset_parents,      # 親セクタのラベル
        values=asset_values,        # セクタごとの値
        branchvalues="total",       # 親が子の階層すべての合計値
    ),
    row=1,
    col=1,
)

sunburst_fig.add_trace(
    go.Sunburst(
        labels=asset_labels,
        parents=asset_parents,
        values=asset_values,
        branchvalues="remainder",   # 子が親とは別の値
    ),
    row=1,
    col=2,
)

sunburst_fig.show()

import plotly.graph_objects as go

companies = ["A社", "B社", "C社", "D社"]
sales_2020 = [900, 700, 200, 100]
sales_2021 = [1400, 980, 450, 380]

go.Figure(
    go.Pie(labels=companies, values=sales_2020),
    layout=go.Layout(title="売上"),  # figureのタイトル
).show()

from plotly.subplots import make_subplots

sales_pie_fig = make_subplots(
    rows=1,
    cols=2,
    specs=[[{"type": "domain"}, {"type": "domain"}]],
    subplot_titles=["2020年の売上", "2021年の売上"],  # サブプロットのタイトル
)
sales_pie_fig.add_trace(
    go.Pie(
        labels=companies,
        values=sales_2020,      # 同じscalegroupでサイズが調整される
        scalegroup="sales",
    ),
    row=1,
    col=1,
)
sales_pie_fig.add_trace(
    go.Pie(
        labels=companies,
        values=sales_2021,
        scalegroup="sales"),    # 同じscalegroupでサイズが調整される
    row=1,
    col=2,
)
sales_pie_fig.show()

import numpy as np
import plotly.graph_objects as go

np.random.seed(7)
area_x = np.arange(15)
area_y1, area_y2 = np.random.rand(2, 15)  # 一様乱数を生成
area_trace = go.Scatter(x=area_x,
                        y=area_y1,
                        name="area1",
                        fill="tozeroy",   # 0からY値まで塗りつぶし,
                        mode="none",      # 線とマーカーを描画しない
                        fillcolor="mediumslateblue"  # 塗りつぶし色
)
area_fig = go.Figure([area_trace])
area_fig.show()

next_area_trace = go.Scatter(x=area_x,
                             y=area_y1 + area_y2,
                             name="area2",
                             fill="tonexty",     # 前のtraceからY値までを塗りつぶし
                             mode="none",
                             fillcolor="lightpink"
)
stacked_area_fig = go.Figure([area_trace, next_area_trace])
stacked_area_fig.show()