Page 546 - Python Data Science Handbook
P. 546
NumPy arrays: multidimensional subarrays, fitting, 408-411
45 kernels and, 411-414
NumPy arrays: one-dimensional subarrays, maximizing the margin, 407-416
44 motivating, 405-420
NumPy vs. Python, 46 simple face detector, 507
Pandas conventions, 114 softening margins, 414-416
sorting arrays, 85-92 surface plots, three-dimensional, 293-298
along rows or columns, 87
basics, 85 T
fast sorting with np.sort and np.argsort, 86 t-distributed stochastic neighbor embedding (t-
k-nearest neighbors example, 88-92 SNE), 456, 472
partitioning, 88 tab completion
source code, accessing, 5 exploring IPython modules with, 6-7
splitting arrays, 49 of object contents, 6
string operations (see vectorized string opera‐ when importing, 7
tions) table, data as, 343
structured arrays, 92-96 target array, 344-345
advanced compound types, 95 term frequency-inverse document frequency
creating, 94 (TF-IDF), 378
record arrays, 96 text, 377
stylesheets (see also annotation of plots)
Bayesian Methods for Hackers, 288 transforms and position of, 270-272
default style, 286 text entry shortcuts, 9
FiveThirtyEight style, 287 three-dimensional plotting
ggplot, 287 contour plots, 292
Matplotlib, 285-290 Möbius strip visualization, 296-298
Seaborn, 289 points and lines, 291
subarrays surface plots, 293-298
as no-copy views, 46 surface triangulations, 295-298
creating copies, 46 wireframes, 293
slicing multidimensional, 45 with Matplotlib, 290-298
slicing one-dimensional, 44 ticks (tick marks)
subplots customizing, 275-282
manual customization, 263-264 fancy formats, 279-281
multiple, 262-268 formatter/locator options, 281
plt.axes() for, 263-264 major and minor, 276
plt.GridSpec() for, 266-268 reducing/increasing number of, 278
plt.subplot() for, 264 Tikhonov regularization, 398
plt.subplots() for, 265 time series
subsets, faceted histograms, 318 bar plots, 321
suffixes keyword, 153 dates and times in Pandas, 191
supervised learning, 332 datetime64, 189
classification task, 333-335 frequency codes, 195
regression task, 335-338 indexing data by timestamps, 192
support vector (defined), 409 native Python dates and times, 189
support vector classifier, 408-411 offsets, 196
support vector machines (SVMs), 405 Pandas, 188-209
advantages/disadvantages, 420 Pandas data structures for, 192-194
face recognition example, 416-420 pd.date_range(), 193
528 | Index

