Data Processing

jupyter with conda:

python -m ipykernel install --user --name 环境名 --display-name “环境名”

1. import packages

import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import icecream
import tqdm as tqdm

2. read files

df_deviceactive = pd.read_csv('./2021_1_data/1_device_active.csv',sep="|",nrows=100000)

3. drawings

plt.scatter(X,Y)
plt.plot(X,Y)
plt.hist(df_active_info['age'])

4. statistics

df_active_info.count() # 显示各列非NAN的数量
```
device_id    2000000
days         2000000
daynum       2000000
gender       1995726
age          1805973
device       2000000
city         2000000
is_vip        247046
topics          8689
dtype: int64
```

df_userinfo['gender'].value_counts() # 统计各种取值的数量
```
0.0    1360810
1.0     634916
Name: gender, dtype: int64
```

df_active_info['age'].dropna() # 去除NAN

5. some snippets

for i,days in tqdm(enumerate(df_deviceactive['days'])): # 迭代