benchmarking

There are a few articles that show that MATLAB prefers column operations than row operations, and that depending on you lay out your data the performance can vary significantly . This is apparently because MATLAB uses a column-major order for representing arrays. I remember reading that Python (NumPy) uses a row-major order. With this, my questions are: Can one expect a similar difference in performance when working with NumPy? If the answer to the above is yes, what would be some examples that highlight this difference ? Like many benchmarks, this really depends on the particulars of the

I'd like to know if there are some benchmark to compare how much cost insert some check constraints on a table of 60 columns where on 20 i'd like to insert a constraints of NotEmpty and on 6 rows NotNull. My case is that i have on my table Empty values and Null values (that in my case means always "no data"). I'd like to unify that data values with just one. That's why I'm thinking to insert NotEmpty constraints on columns, because as i have read null value are not heavy (in byte size) as like empty values (and respect his real meaning). But from other side NotNull Constraint are more deep

How can one measure/benchmark maximum memory usage of a subprocess executed within python? I made a little utility class that demonstrates how to do this with the psutil library : import psutil import subprocess class ProcessTimer: def __init__(self,command): self.command = command self.execution_state = False def execute(self): self.max_vms_memory = 0 self.max_rss_memory = 0 self.t1 = None self.t0 = time.time() self.p = subprocess.Popen(self.command,shell=False) self.execution_state = True def poll(self): if not self.check_execution_state(): return False self.t1 = time.time() try: pp = psutil

I'm doing a benchmark on Cassandra's Reading performance. In the test-setup step I created a cluster with 1 / 2 / 4 ec2-instances and data nodes. I wrote 1 table with 100 million of entries (~3 GB csv-file). Then I launch a Spark application which reads the data into a RDD using the spark-cassandra-connector. However, I thought the behavior should be the following: The more instances Cassandra (same instance amount on Spark) uses, the faster the reads! With the writes everything seems to be correct (~2-times faster if cluster 2-times larger). But: In my benchmark the read is always faster with