# 安裝對應的第三方庫
!pip install pandas  
!pip install numpy
!pip install seaborn
!pip install matplotlib
!pip install sklearn
!pip install tensorflow

#a|T + enter  notebook運行方式
import pandas as pd # 數據處理
import numpy as np # 科學運算
import seaborn as sns # 數據可視化
import matplotlib.pyplot as plt # 數據可視化
import warnings
import warnings
warnings.filterwarnings("ignore")

# 設置圖表與 線格式
plt.rcParams["figure.figsize"] = (10, 10)
plt.rcParams["lines.linewidth"] = 2
plt.style.use("ggplot")
# 讀取數據集
df = pd.read_csv("./DOGE-USD.csv")
df.head() # 查看前5行

df.isnull().sum() # 統計缺失值的總和(sum())
Date         0
Open         5
High         5
Low          5
Close        5
Adj Close    5
Volume       5
dtype: int64
df.duplicated().sum() # 查看重復值
0
# 數據類型 分布基本情況
df.info()
<class "pandas.core.frame.DataFrame">
RangeIndex: 2591 entries, 0 to 2590
Data columns (total 7 columns):
 #   Column     Non-Null Count  Dtype  
---  ------     --------------  -----  
 0   Date       2591 non-null   object 
 1   Open       2586 non-null   float64
 2   High       2586 non-null   float64
 3   Low        2586 non-null   float64
 4   Close      2586 non-null   float64
 5   Adj Close  2586 non-null   float64
 6   Volume     2586 non-null   float64
dtypes: float64(6), object(1)
memory usage: 141.8+ KB
# 轉換 Date的類型
df["Date"] = pd.to_datetime(df.Date, dayfirst=True)
# 索引重置 讓Date時間格式成為 索引  inplace新建對象
df.set_index("Date", inplace=True)
df

df = df.asfreq("d") # 按照天數采集數據
df = df.fillna(method="bfill") # 缺失值填充 下一條數據填充
df

# 開盤價的分布情況
df["Open"].plot(figsize=(12, 8))

# 成交情況
df["Volume"].plot(figsize=(12, 8))

# 投資價值
df["Total Pos"] = df.sum(axis=1)
df["Total Pos"].plot(figsize=(10, 8))

# 當前元素與先前元素的相差百分比
df["Daily Reture"] = df["Total Pos"].pct_change(1)
# 日收益率的平均
df["Daily Reture"].mean()
df["Daily Reture"].plot(kind="kde")

SR = df["Daily Reture"].mean() / df["Daily Reture"].std()
all_plot = df/df.iloc[0]
all_plot.plot(figsize=(24, 16))

df.hist(bins=100, figsize=(12, 6))

# 按照年份進行采樣
df.resample(rule="A").mean()

# 年平均收盤價
df["Open"].resample("A").mean().plot.bar(title="Yearly Mean Closing Price", color=["#b41f7d"])

# 月度
df["Open"].resample("M").mean().plot.bar(figsize=(18, 12), color="red")

# 分別獲取對應時間窗口  6 12 2 均值
df["6-month-SMA"] = df["Open"].rolling(window=6).mean()
df["12-month-SMA"] = df["Open"].rolling(window=12).mean()
df["2-month-SMA"] = df["Open"].rolling(window=2).mean()
df.head(10)

df[["Open", "6-month-SMA", "12-month-SMA", "2-month-SMA"]].plot(figsize=(24, 10))

df[["Open","6-month-SMA"]].plot(figsize=(18,10))

df[["Open","6-month-SMA"]].iloc[:100].plot(figsize=(12,6)).autoscale(axis="x",tight=True)

df["EWMA12"] = df["Open"].ewm(span=14,adjust=True).mean()
df[["Open","EWMA12"]].plot(figsize=(24,12))

df[["Open","EWMA12"]].iloc[:50].plot(figsize=(12,6)).autoscale(axis="x",tight=True)