search

lifeomic / sparkflow

Easy to use library to bring Tensorflow on Apache Spark
MIT License
298 stars 46 forks source link

ord() python3 error #16

Closed huseinzol05 closed 6 years ago

huseinzol05 commented 6 years ago 
/usr/local/lib/python3.5/dist-packages/sparkflow/pipeline_util.py in <listcomp>(.0)
    107         dmp = zlib.compress(dmp)
    108         sc = SparkContext._active_spark_context
--> 109         pylist = [str(ord(d))+ ',' for d in dmp] # convert bytes to string integer list
    110         pylist = [''.join(pylist)]
    111         pylist.append(PysparkObjId._getPyObjId()) # add our id so PysparkPipelineWrapper can id us.

TypeError: ord() expected string of length 1, but int found

When fitted_model.save('simple_dnn')

dmmiller612 commented 6 years ago

Hello and thanks for filing an issue. What version of SparkFlow are you using?

huseinzol05 commented 6 years ago

latest from pip, thank you so much!

FancyXun commented 6 years ago

Remove ord in python3

huseinzol05 commented 6 years ago

yep, thats all i guess

dmmiller612 commented 6 years ago

Yep, I'll get a fix in today.

dmmiller612 commented 6 years ago

Added a fix. I will add the new pip version tonight.

dmmiller612 commented 6 years ago

0.5.0 release is now out in pypi