Da Li (李达)

I tested related code performance on my local machine. ``` # float version # normal pandas def _test_pandas(df): df1 = df[ (df.A > 0.5) & (df.B > 0.5) & (df.C...

I'm still following this issue... We need to refactor/enhance related parts in our jit compiler first... I will sync when ready.

BTW, recently I saw JAX and numba-scipy both contain some scipy.special module code for faster execution speed. Does Mars have interface to choose these execution engine? Or have any schedule...

@ehsantn , yes, we investigated related tools about pandas JIT compiler, and we also think bodo is a wonderful tool for our purpose. However bodo is not open-source, so we...

@ehsantn, currently, we want the single-node version of JIT compiler for Pandas operations, since we use Mars to manage distribution issues (Actually, single-node version of JIT compiler could be integrated...

@kozlov-alexey Hi, I hope to collerabrate with SDC more deeply. My collegues from AntGroup and AliCloud have already focused on [Mars](https://github.com/mars-project/mars) project for several years. U can see Mars' main...

Emm, in my internal test case, it does boost the performance by 10%~20%. My case is heavily using `structref`. Will try to provide a simple benchmark for this.

Recently, I noticed a mem leak in numba NRT if disabling atomicity, (and I believe I use numba in a single thread mode, no concurrent access from different threads or...

close this for now. If needed, please reopen it.

This test config `numba.numba (Linux py310_np123)` seems failed at installing python packages..