42 Dealing with too few experiments

Kapitel in diesem Buch

1. Frontmatter I
2. Acknowledgements VII
3. Contents XIII
4. Introduction and challenge 1
5. Basics 3
6. 1 Getting hands on Python 4
7. 2 Using virtual environments 6
8. 3 Configuring your integrated development environment 9
9. 4 Having a GitHub account 12
10. 5 Creating repositories for dedicated projects 14
11. 6 Synchronizing GitHub desktop 16
12. 7 Knowing basic markdown 19
13. Organization 21
14. 8 Having the overall concept sketch in mind 25
15. 9 Initializing a project with poetry 27
16. 10 Tracking the environment 30
17. 11 Getting your paths right 32
18. 12 Preparing to share 35
19. 13 Writing convenience functions 38
20. 14 Using TOML files for configuration 41
21. 15 Getting used to testing 43
22. Interfacing with common data formats 47
23. 16 Reading Excel files 48
24. 17 Reading text files 51
25. 18 Reading text from Word files 54
26. 19 Reading tables from Word files 57
27. 20 Reading PDF files 59
28. 21 Parsing website contents 61
29. 22 Leveraging regular expressions 64
30. 23 Writing to a database 67
31. 24 Reading from a database 71
32. Planning experiments and/or building on legacy data/information 77
33. 25 Leveraging existing experiments 78
34. 26 Planning experiments 81
35. 27 Using legacy and planned experiments hand in hand 87
36. Collecting experimental data / lab work phase 93
37. 28 Using dedicated modules – use what’s available 94
38. 29 Using dedicated modules – build what’s missing 99
39. Visualization of experimental results 103
40. 30 Simplicity of matplotlib 105
41. 31 Creating a custom matplotlib style 109
42. 32 Convenience of seaborn 112
43. 33 Interactivity of plotly 115
44. 34 Representing multidimensional data 118
45. 35 Representing multidimensional data in a funny way 124
46. Approaching the scientific questions (modeling and recommendation) 131
47. 36 Picking relevant data and information 132
48. 37 Building a model with gplearn 138
49. 38 Playing with the model or “what if” 145
50. 39 Playing with the model or – jupyter notebook 153
51. 40 Playing with the model or – voila 157
52. 41 Playing with the model or – streamlit 160
53. 42 Dealing with too few experiments 166
54. 43 Solving the reverse problem applying multiobjective optimization 173
55. 44 Ensuring the envisioned causality 180
56. Sharing the project 187
57. 45 Building files for distribution 188
58. 46 Pushing to package indices 190
59. 47 Sharing streamlit applications 193
60. Further reading 197
61. 48 Ensuring code styling via black 198
62. 49 Configuring pre-commit 201
63. 50 Building standalone solutions via PyQt 204
64. Concluding remarks 207
65. List of Figures 211
66. Index 215