Question

我有一个Dask图像数据数组。如何将该数组保存到tiff文件的目录中？

Answer 1

理想情况下会有某种select p.pid, p.userID, p.parentID, p.title, p.date, u.userID, u.username, u.loginDate, group_concat(t.tag order by t.tag) tags from postTable as p inner join userTable as u on on p.userID = u.userID inner join tagTable as t on p.pid = t.pid where p.parentID = '0' group by p.pid, p.userID, p.parentID, p.title, p.date, u.userID, u.username, u.loginDate order by p.date limit 10函数。截至2019年10月31日，我找不到一个，但是您将来可能希望观看https://image.dask.org来获得这种功能。

您今天可以使用map_blocks和skimage.io.imread

完成此操作

imsave

。

import dask.array as da
import skimage.io

x = da.random.random((2000, 2000), chunks=(200, 200))  # make a dask array

def save_file(arr, block_info=None):
    """ Save file to foo-x-y.tif, where x and y are block locations """
    filename = "foo-" + "-".join(map(str, block_info[0]["chunk-location"])) + ".tif"
    skimage.io.imsave(filename, arr)
    return arr

x.map_blocks(save_file, dtype=x.dtype).compute()       # call function on every block

如何将Dask阵列另存为一堆图像？

1 个答案: