前段时间在做一些气象预测方面的工作,牵扯到大量的复杂的数据分析与预处理。
该篇文章简述我在用随机森林进行数据分析,计算各类天气参数对于目标参数的贡献度,也就是参数权重大小。
首先引入各个计算工具包
from sklearn.ensemble import RandomForestClassifier
import pandas as pd
import numpy as np
import sys
import csv
接下来对处理好的天气参数csv文件的读取,即x的读取,我所用到的X是 100维的
#读取x值即10x10的风数据值,并将数据转化为numpy的array格式并输入模型
path = sys.path[0] + '/'
file = path + 'weither/data/z100_getdata.csv'
csvfile = open(file, encoding='utf-8')
csvreader = csv.reader(csvfile)
xdata = []
xname = []
timex = []
for x in csvreader:
if x[1] == '925':
timex.append(x[0][:8])
x = [float(x) for x in x[2:]]
xdata.append(x)
print(timex[:10])#该包含time的部分可以省略,是为若有缺失值,填充做的准备
for m in range(100):
xname .append(str('x'+str(m)))
xdata.pop(-1)
xdata = np.array(xdata)
xname = np.array(xname)
#print(xdata[:5],len(xdata),type(xdata),xname[:5])
df = pd.DataFrame(xdata, columns=xname)
df['is_train'] = np.random.uniform(0, 1, len(df)) <= .75#将数据集分为训练集与测试集
接下来读取y值并进行预处理。y值是一维数据,数量要求与x的长度完全一样,yname参数保存数据的分类名
#读取y值即PM2.5的数据值,并将数据转化为numpy的array格式并输入模型
file = path + 'weither/data/sites_2016_getdata.csv'
csvfile = open(file, encoding='utf-8')
csvreader = csv.reader(csvfile)
ydataf = []
yname = []
ydata = []
timey = []
for x in csvreader:
if x[0][0] == '2':
timey.append(x[0])
x = [int(x) for x in x[1:]]
ydataf.extend(x)
print(timey[:10])
cha = set(timex) - set(timey)
print(cha)
i = 0
while i <= len(ydataf)-6:
c = int(sum(ydataf[i:i+6])/240)
if c > 12:
ydata.append(13)
else:
ydata.append(c)
i = i + 6
#print(ydata[:5],len(ydata),type(ydata))
for m in range(len(set(ydata))):
yname .append(str('y'+str(m*40)))
ydata.pop(0)
np.array(ydata)
np.array(yname)
#print(ydata[:5],set(ydata),yname)
接下来是做随机森林,将参数导入,直接调用sklearn里面的RandomForestClassifier即可
df['species'] = pd.Categorical.from_codes(ydata, yname)
#print(df['species'])
df.head()
train, test = df[df['is_train']==True], df[df['is_train']==False]
features = df.columns[:100]
clf = RandomForestClassifier(n_jobs=2)
y, _ = pd.factorize(train['species'])
clf.fit(train[features], y)
preds = np.array(yname)[clf.predict(test[features])]
#print(preds)
pd.crosstab(test['species'], preds, rownames=['actual'], colnames=['preds'])
mmm = clf.feature_importances_
nnn = list(mmm)
nnn = [x*100 for x in nnn]
nnn = [float('%.5f' % x) for x in nnn]
print(nnn)
接下来是对数据的说明:
这是sklearn里面的举例数据集:一条xdata数据有四个值,对应一个y,格式是'numpy.ndarray'格式,但后来我替换成列表格式,不进行numpy转换,似乎也没有出现问题。所用到的数据有xdata\ xname\ ydata\ yname
{'xdata': array([[5.1, 3.5, 1.4, 0.2], [4.9, 3. , 1.4, 0.2], [4.7, 3.2, 1.3, 0.2], [4.6, 3.1, 1.5, 0.2], [5. , 3.6, 1.4, 0.2], [5.4, 3.9, 1.7, 0.4], [4.6, 3.4, 1.4, 0.3], [6. , 3. , 4.8, 1.8], [6.9, 3.1, 5.4, 2.1], [6.7, 3.1, 5.6, 2.4], ...... [6.9, 3.1, 5.1, 2.3], [5.8, 2.7, 5.1, 1.9], [6.8, 3.2, 5.9, 2.3], [6.7, 3.3, 5.7, 2.5], [6.7, 3. , 5.2, 2.3], [6.3, 2.5, 5. , 1.9], [6.5, 3. , 5.2, 2. ], [6.2, 3.4, 5.4, 2.3], [5.9, 3. , 5.1, 1.8]]), 'ydata': array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,......, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,......, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]), 'yname': array(['setosa', 'versicolor', 'virginica'], dtype='