python堆栈内存溢出

浦大志

报错
Traceback (most recent call last):
  File "d:/py/jiwen/total1.py", line 20, in <module>
    Q=0.123*((rsds*60*60*30*24)/(1000*10000))/(4250*4.18)  #Q为光合生产潜力
  File "D:\Python\lib\site-packages\xarray\core\dataarray.py", line 2690, in func
    f(self.variable, other_variable)
  File "D:\Python\lib\site-packages\xarray\core\variable.py", line 2049, in func
    f(self_data, other_data)
MemoryError: Unable to allocate 408. MiB for an array with shape (1032, 180, 288) and data type float64


代码如下
import xarray as xr
import numpy as np
f1= xr.open_dataset(r'D:\py\pr_Amon_FGOALS-f3-L_ssp370_r1i1p1f1_gr_201501-210012.nc')
f2 = xr.open_dataset(r'D:\py\evspsbl\evspsbl_Amon_FGOALS-f3-L_ssp370_r1i1p1f1_gr_201501-210012.nc')
f3 = xr.open_dataset(r'D:\py\rsds\rsds_Amon_FGOALS-f3-L_ssp370_r1i1p1f1_gr_201501-210012.nc')
f4 = xr.open_dataset(r'D:\py\mtas_Amon_FGOALS-f3-L_ssp370_r1i1p1f1_gr_201501-210012.nc')
f=xr.merge([f1,f2['evspsbl'],f3['rsds'],f4['tas']])
pr=f['pr']
eva=f['evspsbl']
rsds=f['rsds']
tas=f['tas']
Q=0.123*((rsds*60*60*30*24)/(1000*10000))/(4250*4.18)  
for i in range(912,920):
    for j in range(92,100):
        for k in range(57,60):
            if tas[i, j, k] >= 283.15:
                pr[i, j, k] = Q[i, j, k] * ((tas[i, j, k] - 273.15) * 30) * (pr[i, j, k] / eva[i, j, k])
            else:
                pr[i,j,k]=0
print(pr)

Lucian

内存不够？可以看看虚拟内存相关

ゼヅ欢チニ

比较好奇，1032*180*288  感觉不大啊，这都能溢出吗？    网上有修改内存的方法，你直接百度下就可以了

Eremite

用dask，官网有，读取时候加上,chunks={"time": 20}这一参数