Ilya: 1 revision imported

2018-07-04T16:28:44Z

1 revision imported

nemenman>Ilya at 19:35, 9 February 2017

2017-02-09T19:35:11Z

New page

{{PHYS212-2017}}

We are playing even more with Python now!

#Python Student Guide, Ch 2.4, 2.5, 2.6, 2.7, and 3.all
#Python Student Guide, Ch 5.1 and 5.8
#Python Student Guide, Ch 4 -- first lab

If you have completed this, convince yourself that memory pre-allocation and vector math make things faster. For this:
# Try to calculate sine of a very long array element by element in a for loop, or on a whole array at once. Which one is faster and why?
# Create a 1000x1000 matrix of numbers from 1 to 1,000,000 three different ways: (i) a single range, sliced as a matrix, (ii) a pre-allocated matrix, filled up by a double for loop, and (iii) 1000 bumpy arrays created as ranges (1...1000), (1001...2000), etc. and stacked on top of each other as a vstack in a for loop. Which one is faster and why?

Finally also check what is faster: using np.arange vs. range when defining a for loop. For this, make a very long for loop that does nothing, and see how long it takes. What is faster and why?

← Older revision	Revision as of 16:28, 4 July 2018
(No difference)

Physics 212, 2017: Lab 4 - Revision history

Ilya: 1 revision imported

nemenman>Ilya at 19:35, 9 February 2017