:orphan: .. _l-notebooks: Notebooks Gallery ================= :ref:`l-notebooks-coverage` .. contents:: :depth: 1 :local: 2016 ---- centrale ++++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/centrale_201606 notebooks/azure_pig notebooks/centrale_201606_sysrec .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/centrale_201606.thumb.png :target: notebooks/centrale_201606.html - :ref:`centrale201606rst` - Présentation à Centrale Paris - Juin 2016. * - .. image:: /notebooks/azure_pig.thumb.png :target: notebooks/azure_pig.html - :ref:`azurepigrst` - Short examples on how to connect to a cluster from a notebook and submit a job (Azure + PIG). * - .. image:: /notebooks/centrale_201606_sysrec.thumb.png :target: notebooks/centrale_201606_sysrec.html - :ref:`centrale201606sysrecrst` - Quelques pistes sur le fonctionnement des moteurs de recommandations sur le web. ensae +++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/kaggle_review_2016 .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/kaggle_review_2016.thumb.png :target: notebooks/kaggle_review_2016.html - :ref:`kagglereview2016rst` - Les gagnants des compétitions Kaggle décrivent parfois leurs solutions sur le blog de Kaggle No Free Hunch. Il y a toujours de bonnes idées à glaner. pydata ++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/10_plotting_libraries notebooks/im_biopython notebooks/js_bokeh notebooks/pyjs_bqplot notebooks/pyjs_brython notebooks/im_cartopy notebooks/big_datashader notebooks/im_ete3 notebooks/gui_geoplotlib notebooks/im_lifelines notebooks/js_lightning_python notebooks/im_matplotlib notebooks/im_missingno notebooks/im_mpl_scatter_density notebooks/js_mpld3 notebooks/im_networkx notebooks/js_plotly notebooks/im_plotnine notebooks/js_pydy_mass_spring_damper notebooks/js_pyecharts notebooks/js_pygal notebooks/js_pythreejs notebooks/im_reportlab notebooks/im_scikit_plot notebooks/im_seaborn notebooks/jsonly_treant notebooks/js_vega notebooks/pyjsc_vispy .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/10_plotting_libraries.thumb.png :target: notebooks/10_plotting_libraries.html - :ref:`10plottinglibrariesrst` - Review of plotting libraries. * - .. image:: /notebooks/im_biopython.thumb.png :target: notebooks/im_biopython.html - :ref:`imbiopythonrst` - The Biopython Project is an international association of developers of freely available Python tools for computational molecular biology. * - .. image:: /notebooks/js_bokeh.thumb.png :target: notebooks/js_bokeh.html - :ref:`jsbokehrst` - bokeh is one if the most mature and complete library using javascript. * - .. image:: /notebooks/pyjs_bqplot.thumb.png :target: notebooks/pyjs_bqplot.html - :ref:`pyjsbqplotrst` - This library is well integrated with Jupyter and will probably stick for a long time. It mixes Python and Javascript. One drawback: you need to run the notebook everytime to get the graph, they don't stay because Jupyter server is sending them, they don't appear in the output. * - .. image:: /notebooks/pyjs_brython.thumb.png :target: notebooks/pyjs_brython.html - :ref:`pyjsbrythonrst` - brython is an implementation of Python in javascript, byrthonmagic makes it available from a notebook. * - .. image:: /notebooks/im_cartopy.thumb.png :target: notebooks/im_cartopy.html - :ref:`imcartopyrst` - *cartopy* aims at drawing maps based on matplotlib. It superimposes a geographical coordinate system on the top of matplotlib's one. It is usually used with modules such as pyproj to handle shapefiles also with shapely or geopandas, fiona, descartes. * - .. image:: /notebooks/big_datashader.thumb.png :target: notebooks/big_datashader.html - :ref:`bigdatashaderrst` - datashader plots huge volume of data. * - .. image:: /notebooks/im_ete3.thumb.png :target: notebooks/im_ete3.html - :ref:`imete3rst` - ete3 draws nice trees. * - .. image:: /notebooks/gui_geoplotlib.thumb.png :target: notebooks/gui_geoplotlib.html - :ref:`guigeoplotlibrst` - geoplotlib implements its own GUI to visualize maps. * - .. image:: /notebooks/im_lifelines.thumb.png :target: notebooks/im_lifelines.html - :ref:`imlifelinesrst` - *lifelines* implements methods and algorithm for life insurance. As many dedicated module, it contains custom graphs built on the top of matplotlib for this module. * - .. image:: /notebooks/js_lightning_python.thumb.png :target: notebooks/js_lightning_python.html - :ref:`jslightningpythonrst` - *lightning-python* is a wrapper for the javascript library lightning. * - .. image:: /notebooks/im_matplotlib.thumb.png :target: notebooks/im_matplotlib.html - :ref:`immatplotlibrst` - matplotlib is the most used to plot. It is the *reference*. * - .. image:: /notebooks/im_missingno.thumb.png :target: notebooks/im_missingno.html - :ref:`immissingnorst` - *missingno* represents missing values in dataframe. * - .. image:: /notebooks/im_mpl_scatter_density.thumb.png :target: notebooks/im_mpl_scatter_density.html - :ref:`immplscatterdensityrst` - mpl-scatter-density speeds up density graph. matplotlib is very slow when it comes to draw millions of points. datashader is one alternative but was meant for zooming/dezooming. This package provides a simple functionality. The example comes the documentation. * - .. image:: /notebooks/js_mpld3.thumb.png :target: notebooks/js_mpld3.html - :ref:`jsmpld3rst` - mpld3 is taking *matplotlib* graphs and converts them into javascript. the support was stopped in Summer 2017. This notebook should be failing at some point is not checked anymore on regular basis. * - .. image:: /notebooks/im_networkx.thumb.png :target: notebooks/im_networkx.html - :ref:`imnetworkxrst` - *networkx* draws networks. It does not work too well on big graphs (< 1000 vertices). * - .. image:: /notebooks/js_plotly.thumb.png :target: notebooks/js_plotly.html - :ref:`jsplotlyrst` - plotly became open source - it was not at the beginning -, it proposes a large gallery of javascript graphs. *plotly* also offers to host dashboards built with plotly. * - .. image:: /notebooks/im_plotnine.thumb.png :target: notebooks/im_plotnine.html - :ref:`implotninerst` - plotnine is an extension of ggplot. The language makes it to compose the data with the layout. I replicate the example from the gallery Two Variable Bar Plot. * - .. image:: /notebooks/js_pydy_mass_spring_damper.thumb.png :target: notebooks/js_pydy_mass_spring_damper.html - :ref:`jspydymassspringdamperrst` - pydy simulates physical systems. * - .. image:: /notebooks/js_pyecharts.thumb.png :target: notebooks/js_pyecharts.html - :ref:`jspyechartsrst` - pyecharts a wrapper for a new library echarts made by Baidu. * - .. image:: /notebooks/js_pygal.thumb.png :target: notebooks/js_pygal.html - :ref:`jspygalrst` - pygal is one if the most mature and complete library using javascript. * - .. image:: /notebooks/js_pythreejs.thumb.png :target: notebooks/js_pythreejs.html - :ref:`jspythreejsrst` - pythreejs allows 3D interactive graphs in a notebook. * - .. image:: /notebooks/im_reportlab.thumb.png :target: notebooks/im_reportlab.html - :ref:`imreportlabrst` - reportlab is the best option if you want to draw graph directly in PDF. Otherwise, code is usually longer with this module compare to *matpotlib* for example. * - .. image:: /notebooks/im_scikit_plot.thumb.png :target: notebooks/im_scikit_plot.html - :ref:`imscikitplotrst` - *scikit-plot* is an extension of matplotlib for datascientist. Proposed graphs are a frequent need when playing with data. * - .. image:: /notebooks/im_seaborn.thumb.png :target: notebooks/im_seaborn.html - :ref:`imseabornrst` - *seaborn* is an extension of matplotlib for statisticians. Graphs are really nice and famous among statisticians. However, it is difficult to draw a graph with weighted observations. * - .. image:: /notebooks/jsonly_treant.thumb.png :target: notebooks/jsonly_treant.html - :ref:`jsonlytreantrst` - treant-js is a javascript library to plot diagram and trees. The goal is to wrap it as a library for python. * - .. image:: /notebooks/js_vega.thumb.png :target: notebooks/js_vega.html - :ref:`jsvegarst` - vega is an extension for the notebook which relies on Vega. * - .. image:: /notebooks/pyjsc_vispy.thumb.png :target: notebooks/pyjsc_vispy.html - :ref:`pyjscvispyrst` - vispy builds graph demanding heavy computation. It does not work well from a notebook. 2017 ---- ensae +++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/2017_1a_ensae_nocture notebooks/kaggle_review_2017 .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/2017_1a_ensae_nocture.thumb.png :target: notebooks/2017_1a_ensae_nocture.html - :ref:`20171aensaenocturerst` - Le data scientiste est devenu populaire. * - .. image:: /notebooks/kaggle_review_2017.thumb.png :target: notebooks/kaggle_review_2017.html - :ref:`kagglereview2017rst` - Les gagnants des compétitions Kaggle décrivent parfois leurs solutions sur le blog de Kaggle No Free Hunch. Il y a toujours de bonnes idées à glaner. meshs +++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/automation_finance_trading .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/automation_finance_trading.thumb.png :target: notebooks/automation_finance_trading.html - :ref:`automationfinancetradingrst` - L'exemple du trading algorithmique. 2018 ---- msexp +++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/onnx_deploy .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/onnx_deploy.thumb.png :target: notebooks/onnx_deploy.html - :ref:`onnxdeployrst` - **Xavier Dupré** - Senior Data Scientist at Microsoft - Computer Science Teacher at ENSAE pyparis +++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/onnx_deploy_pyparis .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/onnx_deploy_pyparis.thumb.png :target: notebooks/onnx_deploy_pyparis.html - :ref:`onnxdeploypyparisrst` - **Xavier Dupré** - Senior Data Scientist at Microsoft - Computer Science Teacher at ENSAE 2019 ---- ensae_api +++++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/sklearn_api .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/sklearn_api.thumb.png :target: notebooks/sklearn_api.html - :ref:`sklearnapirst` - *scikit-learn* est devenu le module incontournable quand il s'agit de machine learning. Cela tient en partie à son API épurée qui permet à quiconque d'implémenter ses propres modèles tout permettant à *scikit-learn* de les manipuler comme s'il s'agissait des siens. sklearn +++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/onnx_sklearn_custom notebooks/onnx_sklearn_consortium .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/onnx_sklearn_custom.thumb.png :target: notebooks/onnx_sklearn_custom.html - :ref:`onnxsklearncustomrst` - The notebook explains how to create a converter for a custom transformer following scikit-learn API. * - .. image:: /notebooks/onnx_sklearn_consortium.thumb.png :target: notebooks/onnx_sklearn_consortium.html - :ref:`onnxsklearnconsortiumrst` - The notebook explains what ONNX is and how it can be used combined with sklearn-onnx and onnxruntime. ONNX is a serialization format for machine learning models. 2020 ---- ensae +++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/introcode .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/introcode.thumb.png :target: notebooks/introcode.html - :ref:`introcoderst` - La programmation est devenue un outil essentiel du datascientist mais pas seulement. Beaucoup d'outils pointus sont open source mais uniquement accessibles à ceux qui savent programmer. Après la mise au point d'un modèle statistique, économique, il se pose souvent la question de la mise à jour fréquente des résultats, c'est à dire leur automatisation via la programmation. 2021 ---- actuaires +++++++++ .. toctree:: :maxdepth: 1 :hidden: notebooks/introprogpy .. list-table:: :header-rows: 0 :widths: 3 5 15 * - .. image:: /notebooks/introprogpy.thumb.png :target: notebooks/introprogpy.html - :ref:`introprogpyrst` - Des années 1970 à la datascience d'aujourd'hui. .. toctree:: :hidden: all_notebooks_coverage