[Dead?] Metro by T-mobile 40刀撸手机指南 [现在好像可以online port in了]

遇到了新的问题,numberbarn的号码失败后一天了还没释放,之前visible的失败了会释放让我再试试

我失败后释放了啊 我一直不能买
我感觉我好累
上班还要搞deal 哎

1 个赞

觉得没释放就是被Metro扣住了,也就是port 成功了。需要打Metro电话释放或下单。参见GV

1 个赞

有人成功过吗?这家客服…

我宁愿号码被吸走也不会打电话给他们客服again

waste of my time

4 个赞

来了来了

根据我自己的dp,一次下了三台后就下不出了,今天三台收到激活后,又可以愉快的下出来了

1 个赞

一共买了几台了?

一台,把间距拉开

请问间隔几天成功概率大一些?

我没有明确几天,至少1-2天,就是想到了就下一台,但是我如果三台在路上,无论怎么下都下不出

1 个赞

不要打电话 灯4-5小时自动释放

我每次打电话都跟他们练口语 他们很nice 能跟我扯1个小时相同的内容

1 个赞

加了

2 个赞

ChatGPT的成功日期data analysis

总结

Advanced Data Analysis Report

1. Time Series Decomposition

The time series decomposition of the gaps into trend, seasonality, and residual components is shown below:

  • Trend: Shows the underlying trend in the gaps over time.
  • Seasonality: Identifies any repeating patterns or cycles.
  • Residuals: Represents the remaining variation after removing trend and seasonality.

2. Autocorrelation Analysis

The autocorrelation function (ACF) plot helps to identify any correlation between the gaps at different lags.

  • The ACF plot does not show significant peaks, indicating a lack of strong autocorrelation.

3. ARIMA Modeling

An ARIMA (AutoRegressive Integrated Moving Average) model was used to forecast the future gaps based on historical data.

  • The red line represents the forecasted gaps for the next 10 periods, showing a continuation of the trend.

请问你有成功从cc port到metro吗?我每次信息卡号填完之后,transfer等好久就会显示失败。

我之前一个号也是一样,似乎无解

CC 的 port out 有问题可以打给 CC 的金牌客服,他们会告诉你 port out request 的问题在哪里,还会主动 offer 做三方通话,但 metro 还是免了,纯纯是在折磨自己和 CC 客服 :yaoming:

我之前遇到的问题是 metro 账户里的 zip code 和 CC 账户里的对不上。

热乎的ACF和PACF

总结

What Do These Plots Tell Us?

  1. Short-term Dependencies: The ACF plot indicates some short-term dependencies (up to lag 2), meaning the gaps between dates are influenced by the previous one or two gaps.
  2. No Strong Long-term Periodicity: Both ACF and PACF plots suggest that there is no strong long-term periodicity or seasonality in the data.
  3. Potential ARIMA Model: The significant PACF value at lag 1 suggests that an AR(1) component is appropriate for the model. The absence of significant higher-order lags suggests that a simple ARIMA(1, d, 0) model might be suitable.

update: K-mean clustering把DP分成了10天以上和以下两组

update2: ChatGPT发现first difference和second difference的plot非常接近,这能推出什么吗?

update3: 尝试了七八个算法后发现楼上的mean就够好了 :troll:

总结

以下是每个模型的最佳参数和评分:

随机森林 (RandomForest)

  • 最佳参数: {'max_depth': None, 'min_samples_split': 2, 'n_estimators': 100}
  • 最佳评分: 0.67

支持向量机 (SVM)

  • 最佳参数: {'C': 0.1, 'gamma': 1, 'kernel': 'linear'}
  • 最佳评分: 0.67

决策树 (DecisionTree)

  • 最佳参数: {'max_depth': 10, 'min_samples_leaf': 1, 'min_samples_split': 2}
  • 最佳评分: 1.00

梯度提升 (GradientBoosting)

  • 最佳参数: {'learning_rate': 0.1, 'max_depth': 3, 'n_estimators': 100}
  • 最佳评分: 0.83

XGBoost

  • 交叉验证评分: [0.67, 0.83, 0.83, 0.67, 0.83]
  • 平均交叉验证评分: 0.77
总结

多模型结合后的预测结果

我们结合了随机森林、决策树和梯度提升模型的预测结果,以下是综合分析后的结果:

未来数据预测结果(2024-06-15至2024-07-02)

     Date  RandomForest  DecisionTree  GradientBoosting  Consensus

0 2024-06-15 2.000000 2.000000 0.0 2.0
1 2024-06-16 4.000000 4.000000 0.0 4.0
2 2024-06-17 5.000000 5.000000 0.0 5.0
3 2024-06-18 6.000000 6.000000 1.0 6.0
4 2024-06-19 6.000000 6.000000 2.0 6.0
5 2024-06-20 2.000000 2.000000 1.0 2.0
6 2024-06-21 5.000000 5.000000 2.0 5.0
7 2024-06-22 4.000000 4.000000 1.0 4.0
8 2024-06-23 1.000000 1.000000 2.0 1.0
9 2024-06-24 5.000000 5.000000 0.0 5.0
10 2024-06-25 3.000000 3.000000 0.0 3.0
11 2024-06-26 3.000000 3.000000 0.0 3.0
12 2024-06-27 10.000000 10.000000 0.0 10.0
13 2024-06-28 5.277778 5.277778 1.0 5.3
14 2024-06-29 1.000000 1.000000 0.0 1.0
15 2024-06-30 5.277778 5.277778 0.0 5.3
16 2024-07-01 6.000000 6.000000 1.0 6.0
17 2024-07-02 6.000000 6.000000 1.0 6.0

好消息是新手机到了,坏消息是又没给sim卡,不能add a line了,我怀疑他们故意的