clustering/wind-fft-clustering/data_add.py

# -*- coding: utf-8 -*-


import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from sklearn.preprocessing import MinMaxScaler
import os


# cluster_power路径位置
root_path = "../data-process/data/"
# 1、2类平均机头风速，总平均机头风速，nwp风速，实际功率
add_cols = ["C_WS_1", "C_WS_2", "C_WS_ALL",
            "C_WS100", "C_WS170", "power", "C_REAL_VALUE"]


# 处理几个表的数据，拼接在一起，得到上述列
def data_process():
    id1 = [142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155,
           156, 157, 158, 159, 160, 161]
    id2 = [102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115,
           116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129,
           130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141]


    df_power = pd.read_csv(root_path + "power.csv")
    df_nwp = pd.read_csv(root_path + "NWP.csv",
                         usecols=["C_TIME", "C_WS100", "C_WS170"])
    # df_nwp_power = pd.merge(df_power, df_nwp, on="C_TIME", how="inner")

    turbine_path = root_path + "output_filtered_csv_files/"
    df_turbine = pd.read_csv(
        turbine_path + "turbine-102.csv", usecols=["C_TIME"])
    df_turbine["C_WS_1"] = [0] * len(df_turbine)
    df_turbine["C_WS_2"] = [0] * len(df_turbine)
    df_turbine["C_WS_ALL"] = [0] * len(df_turbine)
    df_turbine["power"] = [0] * len(df_turbine)

    for ids in id1:
        df_temp = pd.read_csv(turbine_path + f"turbine-{ids}.csv")
        # if len(df_temp) != len(df_turbine):
        #     print("false")
        df_turbine["C_WS_1"] += df_temp["C_WS"]
        df_turbine["C_WS_ALL"] += df_temp["C_WS"]
        df_turbine["power"] += df_temp["C_ACTIVE_POWER"]

    df_turbine["C_WS_1"] /= len(id1)

    for ids in id2:
        df_temp = pd.read_csv(turbine_path + f"turbine-{ids}.csv")
        # if len(df_temp) != len(df_turbine):
        #     print("false")
        df_turbine["C_WS_2"] += df_temp["C_WS"]
        df_turbine["C_WS_ALL"] += df_temp["C_WS"]
        df_turbine["power"] += df_temp["C_ACTIVE_POWER"]

    df_turbine["C_WS_2"] /= len(id2)
    df_turbine["C_WS_ALL"] /= (len(id1) + len(id2))
    df_turbine["power"] /= (len(id1) + len(id2))

    df_all = pd.concat([df_power.set_index("C_TIME"), df_nwp.set_index("C_TIME"),
                        df_turbine.set_index("C_TIME")], axis=1, join="inner").reset_index()
    df_all = df_all.reindex(columns=["C_TIME"] + add_cols)
    # df_all.drop(columns="power", inplace=True)
    df_all.to_csv(root_path + "df_all.csv", index=False)


# data_process()


# 在cluster_data.csv中新增若干列（add_cols），得到cluster_data_1.csv
def data_add(dirname, filename):
    df_temp = pd.read_csv(dirname + filename)
    df_all = pd.read_csv(root_path + "df_all.csv")
    df = pd.merge(df_all, df_temp, on="C_TIME", how="inner")
    df = df.reindex(columns=["C_TIME", "power_1",
                    "power_2"] + add_cols + ["SUM"])
    df.to_csv(dirname + "cluster_data_1.csv", index=False)


# 画随时间变化的曲线
def show_curve(dirname, filename, series1, series2):
    df = pd.read_csv(dirname + filename)

    cols = df.columns[1:]
    scaler = MinMaxScaler()
    # 最大最小归一化
    df[cols] = scaler.fit_transform(df[cols])

    c_time = pd.to_datetime(df["C_TIME"])

    plt.figure(figsize=(12, 8), dpi=100)

    plt.plot(c_time, df[series1], label=series1)
    plt.plot(c_time, df[series2], label=series2)

    plt.legend()
    date_format = mdates.DateFormatter('%Y-%m-%d %H:%M')
    plt.gca().xaxis.set_major_formatter(date_format)
    plt.xticks(rotation=30)

    plt.show()
    plt.savefig(dirname + "curve_" + series1 + "_" + series2 + ".png")
    plt.close()


# 画s型曲线
def show_scatter(dirname, filename, series1, series2, series3):
    df = pd.read_csv(dirname + filename)

    cols = df.columns[1:]
    scaler = MinMaxScaler()
    # 最大最小归一化
    # df[cols] = scaler.fit_transform(df[cols])

    plt.figure(figsize=(10, 8), dpi=100)
    point_size = 10
    plt.scatter(df[series1], df[series3], label=series1, s=point_size)
    plt.scatter(df[series2], df[series3], label=series2, s=point_size)
    plt.xlabel(series1 + " / " + series2)
    plt.ylabel(series3)

    plt.legend()
    plt.show()
    plt.savefig(dirname + "scatter_" + series1 +
                "_" + series2 + "_" + series3 + ".png")
    plt.close()


# %%
if __name__ == "__main__":
    cluster_path = root_path + "cluster_power/"
    # 新增数据
    data_add(cluster_path, "cluster_data.csv")

    for root, dirs, files in os.walk(cluster_path):
        for sub_dir in dirs:
            subdir_path = os.path.join(root, sub_dir)
            # print(subdir_path)
            # file_path = os.path.join(subdir_path, "cluster_data.csv")
            # print(file_path)

            data_add(subdir_path + '/', "cluster_data.csv")

    # %% 画曲线图
    show_curve(cluster_path, "cluster_data_1.csv", "SUM", "C_WS_ALL")
    for root, dirs, files in os.walk(cluster_path):
        for sub_dir in dirs:
            subdir_path = os.path.join(root, sub_dir)
            show_curve(subdir_path + "/",
                       "cluster_data_1.csv", "SUM", "C_WS_ALL")
            # show_curve(subdir_path + "/", "cluster_data_1.csv", "power_1", "C_WS_1")
            # show_curve(subdir_path + "/", "cluster_data_1.csv", "power_2", "C_WS_2")

    # %% 画散点图（s型曲线）
    show_scatter(cluster_path, "cluster_data_1.csv",
                 "C_WS_ALL", "C_WS100", "SUM")
    for root, dirs, files in os.walk(cluster_path):
        for sub_dir in dirs:
            subdir_path = os.path.join(root, sub_dir)
            show_scatter(subdir_path + "/", "cluster_data_1.csv",
                         "C_WS_ALL", "C_WS100", "SUM")