<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[sphinxrascal168的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/sphinxrascal168</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; sphinxrascal168]]></copyright><item><title><![CDATA[诗词偶得，胡思乱想]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/140496968</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/140496968</guid><author>sphinxrascal168</author><pubDate>Wed, 17 Jul 2024 15:42:01 +0800</pubDate><description><![CDATA[群山披色，斑驳雄奇。日渐沉于群山之后，光亦逐敛收云层之中。众兄弟起身，迎风而立，同归营地，众也四散，皆无言以述，盖其美无言可述之也。月亮之谷，落日时分，有客徐来，皆寻石而坐。三五成群，静默无语。偶倾，日自云幕而降，金光四溢，千里如血，汇丰手机银行 8211**平安银行密码360***]]></description><category></category></item><item><title><![CDATA[【无标题】放大放大放大]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/134259166</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/134259166</guid><author>sphinxrascal168</author><pubDate>Tue, 07 Nov 2023 09:39:06 +0800</pubDate><description><![CDATA[我完全不知道你现在消耗多大，法院给到你的6万多，加上我转的和你自己原有的，好像不足3个月就已经不够用了。争论不休的的根源是，我们都对自己应该做什么没有清晰的认识，都是说了自己干了一件具体的事情，就好像已经干了所有的事情。3、家庭作为一个最小的组织，是要靠双方共同努力的，不是一个人要给对方什么生活，一个人要给对方找个工作，如果没有，那就可以自己随便来，这样的说法，是自贬身价的。1、首先是对于责任的共识。别人的热闹都是别人的，别人最多要你给个围观，我们没有能力，不能给别人带来收益，是没人要真和你一起合作的。]]></description><category></category></item><item><title><![CDATA[循环for处理数据]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/127764166</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/127764166</guid><author>sphinxrascal168</author><pubDate>Wed, 09 Nov 2022 10:04:08 +0800</pubDate><description><![CDATA[一种循环取数的方法]]></description><category></category></item><item><title><![CDATA[把两个毫不相关dataframe中的字段合并在一起]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123945495</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123945495</guid><author>sphinxrascal168</author><pubDate>Sun, 03 Apr 2022 21:54:15 +0800</pubDate><description><![CDATA[df_target = pd.DataFrame(df_pct[(df_pct.ts_code == code)].sort_values(by=['date'],ascending=False))
  df_xiang = pd.DataFrame(df_target['ts_code'].head(1))
  df_hebing = pd.merge(df_pct,pivot_df,on=['date'],how='inner')
  df_hebing = df_hebing[:-...]]></description><category></category></item><item><title><![CDATA[组内排序 python和sql]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123439405</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123439405</guid><author>sphinxrascal168</author><pubDate>Sat, 12 Mar 2022 10:23:07 +0800</pubDate><description><![CDATA[#生成rownum
df1 = df1.sort_values(by=['ts_code','trade_date'],ascending=True)
df1['rownum'] = df1['trade_date'].groupby(df1['ts_code']).rank(ascending=False)#tscode组内排序,按日期降序





SELECT
	t_time,
	code,
	name,
	CL,
	row_number(...]]></description><category></category></item><item><title><![CDATA[卫裤选择标准]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123242180</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123242180</guid><author>sphinxrascal168</author><pubDate>Wed, 02 Mar 2022 22:30:52 +0800</pubDate><description><![CDATA[卫裤还是85%的棉的好，不起球
]]></description><category></category></item><item><title><![CDATA[打印机设置]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123174056</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/123174056</guid><author>sphinxrascal168</author><pubDate>Mon, 28 Feb 2022 09:11:08 +0800</pubDate><description><![CDATA[1、从控制面板 进入 插看设备和打印机





2、选中一个打印机，查看打印服务器属性，在安全选项卡中，把管理服务器权限的√去掉





3、则此时，端口和驱动程序下的添加、删除、配置等按钮变灰，无法安装打印机，重新把√选中则可以恢复




...]]></description><category></category></item><item><title><![CDATA[时间序列 fbprophet]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122976391</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122976391</guid><author>sphinxrascal168</author><pubDate>Thu, 17 Feb 2022 09:35:11 +0800</pubDate><description><![CDATA[第一天占比多少



总额



趋势：





Profit 月效应



一、1月总消费金额：

方法：2年的数据，增长率为主。





二、计算趋势

H2o：输入18年每一天的消费金额，日期，日期的属性，周几，是否节假日，间隔几天 ，得到1月每天的消费金额，你会计算占比，然后来调整。

Facebookprofit模型：输入变量，有节假日参数， 输出的结果很准，月末的几天就会高一些。





三、不是客户级的，可用lightgbm来计算客户级的，但是都误差大， 合起来。
...]]></description><category></category></item><item><title><![CDATA[lightgbm调参后人工修改的结果记录20220216]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122971435</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122971435</guid><author>sphinxrascal168</author><pubDate>Wed, 16 Feb 2022 21:10:50 +0800</pubDate><description><![CDATA[761688

clf = LGBMClassifier(reg_alpha=0.15,reg_lambda=10,num_leaves=60,max_depth=7,n_estimators=1500,
          learning_rate=0.01,bagging_fraction=0.7,bagging_freq=10,max_bin=150,
          min_data_in_leaf=30,feature_fraction=0.7,r...]]></description><category></category></item><item><title><![CDATA[布林线的python定义]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122874151</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122874151</guid><author>sphinxrascal168</author><pubDate>Fri, 11 Feb 2022 09:50:52 +0800</pubDate><description><![CDATA[#布林线相关指标,shift(-1),向上移动1格，也就当前行变成下一行，在这里就是未来的某一天，+1就是向下移动，就是过去的一天
df1['upper_BBANDS'],df1['mid_BBANDS'],df1['lower_BBANDS'] = ta.BBANDS(df1['close'].values,timeperiod=5,nbdevup=2,nbdevdn=2,matype=0)

df1['mid_BBANDS_xielv'] =ta.LINEARREG_SLOPE(df1['mid_BBA]]></description><category></category></item><item><title><![CDATA[shapley]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122870421</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122870421</guid><author>sphinxrascal168</author><pubDate>Thu, 10 Feb 2022 22:36:32 +0800</pubDate><description><![CDATA[import lightgbm as lgb
import shap
model = lgb.LGBMClassifier().fit(X,y)
explainer = shap.Explainer(model)
shap_value = explainer(X)

shap.plots.waterfall(shap_value[100],max_display=40)



shap.plots.bar(shap_value,max_display=30)
]]></description><category></category></item><item><title><![CDATA[完善一些内容]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122719317</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122719317</guid><author>sphinxrascal168</author><pubDate>Thu, 27 Jan 2022 16:05:07 +0800</pubDate><description><![CDATA[#################开始训练模型###########################

train_data = df4.loc[(df1['rownum_W']&gt;=sst) &amp; (df1['rownum_W']&lt;=101)]

train_data = train_data.fillna(0)

#处理以下target的字段类型，不然后面有问题
train_data['target'] = train_data['target'].astype('int')
data_]]></description><category></category></item><item><title><![CDATA[一些错误的修复笔记]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122236970</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122236970</guid><author>sphinxrascal168</author><pubDate>Thu, 30 Dec 2021 15:13:08 +0800</pubDate><description><![CDATA[第10行

df1['c_ma30'] = TA.SMA(df1,30,column='close')#原来的跟21重合了，搞错了

第13

#21日均线的斜率是否大于0，来判断是否盘整等
df1['ma21xielv_d0_flag'] = np.where(df1['ma21xielv']&gt;=0,1,0)
df1['d21_0_count'] = ta.SUM(df1['ma21xielv_d0_flag'], timeperiod=15)



#盘平df1['d21_0_count'] = ]]></description><category></category></item><item><title><![CDATA[学习笔记 select的用法]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122211369</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122211369</guid><author>sphinxrascal168</author><pubDate>Wed, 29 Dec 2021 11:23:56 +0800</pubDate><description><![CDATA[#####对当日涨跌进行分段
df1['pct_chg_cate'] = np.select([((df1['pct_chg']&gt;=4.9999)),
                 ((df1['pct_chg'] &lt;4.9999) &amp; (df1['pct_chg']&gt;=1.5001)),
                ((df1['pct_chg'] &lt;(1.5001)) &amp; (df1['pc...]]></description><category></category></item><item><title><![CDATA[一些更正的东西]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122184825</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122184825</guid><author>sphinxrascal168</author><pubDate>Tue, 28 Dec 2021 09:47:24 +0800</pubDate><description><![CDATA[第13行

#是否长上影，绝对长度和相对长度都有要求
df1['K_l_up_flag'] = np.where((df1['shiti_up']&gt;50) &amp; (df1['shiti_up_len']&gt;3),1,0)
#是否长下影
df1['K_l_low_flag'] = np.where((df1['shiti_low']&gt;50) &amp; (df1['shiti_low_len']&gt;3),1,0)
#是否上下影都长
df1['K_l_lu_flag'] = np.wh]]></description><category></category></item><item><title><![CDATA[python盘整之路]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122176454</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/122176454</guid><author>sphinxrascal168</author><pubDate>Mon, 27 Dec 2021 17:13:00 +0800</pubDate><description><![CDATA[#以今天的上线浮动12%，看是否盘整
df1['c_up_12per'] = df1['close'].shift(14)*1.12
df1['c_down_12per'] = df1['close'].shift(14)*0.88
df1['c_shitf1_nei_flag'] = np.where((df1['close'].shift(1)&gt; df1['c_down_12per']) &amp; (df1['close'].shift(1)&lt; df1['c_up_12per']),1..]]></description><category></category></item><item><title><![CDATA[var——ok]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119791915</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119791915</guid><author>sphinxrascal168</author><pubDate>Thu, 19 Aug 2021 00:07:57 +0800</pubDate><description><![CDATA[var_ok = list(set(['shizi_panzheng_count','zhenfu_count_W','l30_c_xielv_W','l30_c_xielv_count_W','h30_c_xielv_count_W','RSI_2_W','bian_2_UP_count_W','yuanhu_count_W','h30_c_xielv_W','suoliang_flag_count_W','fangliang_flag_count_W','xie_l_up_count_W','hengc]]></description><category></category></item><item><title><![CDATA[特征选取的boruta]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119495554</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119495554</guid><author>sphinxrascal168</author><pubDate>Sun, 08 Aug 2021 00:20:02 +0800</pubDate><description><![CDATA[# X=X.values
# y=y.values
#更改为bor特征选择器
# import sklearn
# from lightgbm.sklearn import LGBMClassifier
# from sklearn.ensemble import RandomForestClassifier
# from boruta import BorutaPy

# rf = LGBMClassifier(num_leaves=60,max_depth=7,n_estimators=1200,lea]]></description><category></category></item><item><title><![CDATA[mysql启动错误10061]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119490931</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119490931</guid><author>sphinxrascal168</author><pubDate>Sat, 07 Aug 2021 18:40:50 +0800</pubDate><description><![CDATA[原因： 没有启动MySQL服务------------更多是删除某个文件库表时导致的错误
解决方案：

以管理员身份打开cmd命令行窗口，使用cd命令将目录切换为你MySQL的bin目录下，输入 net start mysql 启动MySQL服务即可解决此问题！！！
电脑中搜索服务找到MySQL服务，启动它！
管理员身份命令行窗口打开Windows快捷键： Win + X + A

如果找不到MySQL服务，需要安装MySQL服务。解决的方法是用管理员的方式打开mysql安装目录bin。执行mysqld.e]]></description><category></category></item><item><title><![CDATA[一些股票代码的留存]]></title><link>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119299666</link><guid>https://blog.youkuaiyun.com/sphinxrascal168/article/details/119299666</guid><author>sphinxrascal168</author><pubDate>Sun, 01 Aug 2021 16:28:00 +0800</pubDate><description><![CDATA[#提前缩小数据量
df_t = df1.loc[(df1['rownum_W'] &lt;= (sst+25))]
t = df_t.groupby(['rownum_W'],as_index=False).ts_code.count()
row_liu = t['rownum_W'].loc[t.ts_code&gt;12000]

df1 = pd.DataFrame()
for row in row_liu:
  df_for = df_t[df_t.rownum_W==row]
  df1 ...]]></description><category></category></item></channel></rss>