BaiJian
BaiJian
See JIRA: https://issues.apache.org/jira/browse/IOTDB-4187
The scenario is as issue#18 said.
In some situation, we may send some auth info like 'access-token' to the server in the header. However the invoke_shell() function doesn't allow developers adding their headers.
This PR adds a univeral function `replace_subgraph` to `OptimizationRule` for later use. The unit test cases are added as well. Fixes #3352 ## Check code requirements - [x] tests added...
**Describe the bug** Series with multi index accessing data by index error, but pandas works well. **To Reproduce** ```python import pandas as pd arrays = [["m1", "m2", "m3"], ["n1", "n2",...
The `Series` doesn't support attribute access, but the [pandas'](https://pandas.pydata.org/pandas-docs/version/1.5/user_guide/indexing.html#attribute-access) can. **To Reproduce** ``` import mars.dataframe as md s1 = md.Series([1, 2, 3], index=list("abc")) print(s1.b) >>> AttributeError: 'SeriesData' object has no...
Currently in OptimizationRule, there're only collapsable_predecessors adding/removing and node replacing functions to handle graph mutation. However, some new rules may need to replace a piece of subgraph by adding/removing nodes...