pytimetk 0.3.0

docs/_site/changelog-news.html

@@ -332,7 +332,7 @@
     <h2 id="toc-title">On this page</h2>
 
   <ul>
-  <li><a href="#pytimetk-0.2.1.9000-development-version" id="toc-pytimetk-0.2.1.9000-development-version" class="nav-link active" data-scroll-target="#pytimetk-0.2.1.9000-development-version">pytimetk 0.2.1.9000 (Development Version)</a>
+  <li><a href="#pytimetk-0.3.0" id="toc-pytimetk-0.3.0" class="nav-link active" data-scroll-target="#pytimetk-0.3.0">pytimetk 0.3.0</a>
   <ul class="collapse">
   <li><a href="#correlation-funnel" id="toc-correlation-funnel" class="nav-link" data-scroll-target="#correlation-funnel">Correlation Funnel</a></li>
   <li><a href="#core" id="toc-core" class="nav-link" data-scroll-target="#core">Core:</a></li>
@@ -381,8 +381,8 @@ <h1 class="title">Changelog for PyTimeTK</h1>
 
 </header>
 
-<section id="pytimetk-0.2.1.9000-development-version" class="level1">
-<h1>pytimetk 0.2.1.9000 (Development Version)</h1>
+<section id="pytimetk-0.3.0" class="level1">
+<h1>pytimetk 0.3.0</h1>
 <section id="correlation-funnel" class="level3">
 <h3 class="anchored" data-anchor-id="correlation-funnel">Correlation Funnel</h3>
 <p>The R package <code>correlationfunnel</code> has been ported inside <code>pytimetk</code>:</p>

docs/_site/search.json

@@ -1642,7 +1642,7 @@
     "href": "changelog-news.html",
     "title": "Changelog for PyTimeTK",
     "section": "",
-    "text": "pytimetk 0.2.1.9000 (Development Version)\n\nCorrelation Funnel\nThe R package correlationfunnel has been ported inside pytimetk:\n\nbinarize()\ncorrelate()\nplot_correlation_funnel()\n\n\n\nCore:\n\nfilter_by_time() - Filtering with time-based strings\n\n\n\nFeature Engineering:\n\naugment_diffs() - Can now add differenced columns\naugment_fourier() - Can now add fourier features.\n\n\n\nFinance Module:\n\naugment_cmo(): Chande Momentum Oscillator (CMO)\n\n\n\nNew Polars Backends:\n\naugment_diffs()\naugment_fourier()\naugment_cmo()\n\n\n\nGeneral\n\nMake memory reduction optional #275\n\n\n\n\npytimetk 0.2.1\nBugfix - Issue with augment_rolling(engine='pandas') and augment_expanding(engine='pandas') with concatinating rolled/expanded calc’s to the correct group\n\n\npytimetk 0.2.0\n\nAnomaly Detection\n\nanomalize(): A new scalable function for detecting time series anomalies (outliers)\nplot_anomalies(): A scalable visualization tool for inspecting anomalies in time series data.\nplot_anomalies_decomp: A scalable visualization tool for inspecting the observed, seasonal, trend, and remainder decomposition, which are useful for telling you whether or not anomalies are being detected to your preference.\nplot_anomalies_cleaned(): A scalable visualization tool for showing the before and after transformation for the cleaned vs uncleaned anomalies.\n\n\n\nNew Functions:\n\napply_by_time(): For complex apply-style aggregations by time.\naugment_rolling_apply(): For complex rolling operations using apply-style data frame functions.\n\naugment_expanding(): For expanding calculations with single-column functions (e.g. mean).\naugment_expanding_apply(): For complex expanding operations with apply-style data frame functions.\naugment_hilbert(): Hilbert features for signal processing.\naugment_wavelet(): Wavelet transform features.\nget_frequency(): Infer a pandas-like frequency. More robust than pandas.infer_freq.\nget_seasonal_frequency(): Infer the pandas-like seasonal frequency (periodicity) for the time series.\nget_trend_frequency(): Infer the pandas-like trend for the time series.\n\n\n\nNew Finance Module\nMore coming soon.\n\naugment_ewm(): Exponentially weighted augmentation\n\n\n\nSpeed Improvements\n\nPolars Engines:\n\nsummarize_by_time(): Gains a polars engine.\n\n3X to 10X speed improvements.\n\naugment_lags() and augment_leads(): Gains a polars engine. Speed improvements increase with number of lags/leads.\n\n6.5X speed improvement with 100 lags.\n\naugment_rolling(): Gains a polars engine. 10X speed improvement.\naugment_expanding(): Gains a polars engine.\naugment_timeseries_signature(): Gains a polars engine. 3X speed improvement.\naugment_holiday_signature(): Gains a polars engine.\n\n\n\nParallel Processing and Vectorized Optimizations:\n\npad_by_time(): Complete overhaul. Uses Cartesian Product (Vectorization) to enhance the speed. 1000s of time series can now be padded in seconds.\n\nIndependent review: Time went from over 90 minutes to 13 seconds for a 500X speedup on 10M rows.\n\nfuture_frame(): Complete overhaul. Uses vectorization when possible. Grouped parallel processing. Set threads = -1 to use all cores.\n\nIndependent Review: Time went from 11 minutes to 2.5 minutes for a 4.4X speedup on 10M rows\n\nts_features: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\nts_summary: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\nanomalize: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\naugment_rolling() and augment_rolling_apply(): Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\n\n\n\n\nHelpful Utilities:\n\nparallel_apply: Mimics the pandas apply() function with concurrent futures.\nprogress_apply: Adds a progress bar to pandas apply()\nglimpse(): Mimics tidyverse (tibble) glimpse function\n\n\n\nNew Data Sets:\n\nexpedia: Expedia hotel searches time series data set\n\n\n\n3 New Applied Tutorials:\n\nSales Analysis Tutorial\nFinance Analysis Tutorial\nDemand Forecasting Tutorial\nAnomaly Detection Tutorial\n\n\n\nFinal Deprecations:\n\nsummarize_by_time(): kind = \"period\". This was removed for consistency with pytimetk. “timestamp” is the default.\naugment_rolling(): use_independent_variables. This is replaced by augment_rolling_apply().\n\n\n\n\npytimetk 0.1.0 (2023-10-02)\n\nAbout the Initial release.\nThis release includes the following features:\n\nA workhorse plotting function called plot_timeseries() 💪\nThree (3) data wrangling functions that will simplify 90% of time series tasks 🙏\nFive (5) “augmentor” functions: These add hundreds of features to time series to help in predictive tasks 🧠\nTwo (2) time series feature summarizes: identify key aspects of your time series 🔍\nNine (9) pandas series and DatetimeIndex helpers (work more easily with these timestamp data structures) ⏲\nFour (4) date utility functions that fill in missing function gaps in pandas 🐼\nTwo (2) Visualization utilities to help you customize your visualizations and make them look MORE professional 📈\nTwo (2) Pandas helpers that help clean up and understand pandas data frames with time series 🎇\nTwelve (12) time series datasets that you can practice PyTimeTK time series analysis on 🔢\n\n\n\nThe PyTimeTK website comes with:\n\nTwo (2) Getting started tutorials\nFive (5) Guides covering common tasks\nComing Soon: Applied Tutorials in Sales, Finance, Demand Forecasting, Anomaly Detection, and more."
+    "text": "pytimetk 0.3.0\n\nCorrelation Funnel\nThe R package correlationfunnel has been ported inside pytimetk:\n\nbinarize()\ncorrelate()\nplot_correlation_funnel()\n\n\n\nCore:\n\nfilter_by_time() - Filtering with time-based strings\n\n\n\nFeature Engineering:\n\naugment_diffs() - Can now add differenced columns\naugment_fourier() - Can now add fourier features.\n\n\n\nFinance Module:\n\naugment_cmo(): Chande Momentum Oscillator (CMO)\n\n\n\nNew Polars Backends:\n\naugment_diffs()\naugment_fourier()\naugment_cmo()\n\n\n\nGeneral\n\nMake memory reduction optional #275\n\n\n\n\npytimetk 0.2.1\nBugfix - Issue with augment_rolling(engine='pandas') and augment_expanding(engine='pandas') with concatinating rolled/expanded calc’s to the correct group\n\n\npytimetk 0.2.0\n\nAnomaly Detection\n\nanomalize(): A new scalable function for detecting time series anomalies (outliers)\nplot_anomalies(): A scalable visualization tool for inspecting anomalies in time series data.\nplot_anomalies_decomp: A scalable visualization tool for inspecting the observed, seasonal, trend, and remainder decomposition, which are useful for telling you whether or not anomalies are being detected to your preference.\nplot_anomalies_cleaned(): A scalable visualization tool for showing the before and after transformation for the cleaned vs uncleaned anomalies.\n\n\n\nNew Functions:\n\napply_by_time(): For complex apply-style aggregations by time.\naugment_rolling_apply(): For complex rolling operations using apply-style data frame functions.\n\naugment_expanding(): For expanding calculations with single-column functions (e.g. mean).\naugment_expanding_apply(): For complex expanding operations with apply-style data frame functions.\naugment_hilbert(): Hilbert features for signal processing.\naugment_wavelet(): Wavelet transform features.\nget_frequency(): Infer a pandas-like frequency. More robust than pandas.infer_freq.\nget_seasonal_frequency(): Infer the pandas-like seasonal frequency (periodicity) for the time series.\nget_trend_frequency(): Infer the pandas-like trend for the time series.\n\n\n\nNew Finance Module\nMore coming soon.\n\naugment_ewm(): Exponentially weighted augmentation\n\n\n\nSpeed Improvements\n\nPolars Engines:\n\nsummarize_by_time(): Gains a polars engine.\n\n3X to 10X speed improvements.\n\naugment_lags() and augment_leads(): Gains a polars engine. Speed improvements increase with number of lags/leads.\n\n6.5X speed improvement with 100 lags.\n\naugment_rolling(): Gains a polars engine. 10X speed improvement.\naugment_expanding(): Gains a polars engine.\naugment_timeseries_signature(): Gains a polars engine. 3X speed improvement.\naugment_holiday_signature(): Gains a polars engine.\n\n\n\nParallel Processing and Vectorized Optimizations:\n\npad_by_time(): Complete overhaul. Uses Cartesian Product (Vectorization) to enhance the speed. 1000s of time series can now be padded in seconds.\n\nIndependent review: Time went from over 90 minutes to 13 seconds for a 500X speedup on 10M rows.\n\nfuture_frame(): Complete overhaul. Uses vectorization when possible. Grouped parallel processing. Set threads = -1 to use all cores.\n\nIndependent Review: Time went from 11 minutes to 2.5 minutes for a 4.4X speedup on 10M rows\n\nts_features: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\nts_summary: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\nanomalize: Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\naugment_rolling() and augment_rolling_apply(): Uses concurrent futures to parallelize tasks. Set threads = -1 to use all cores.\n\n\n\n\nHelpful Utilities:\n\nparallel_apply: Mimics the pandas apply() function with concurrent futures.\nprogress_apply: Adds a progress bar to pandas apply()\nglimpse(): Mimics tidyverse (tibble) glimpse function\n\n\n\nNew Data Sets:\n\nexpedia: Expedia hotel searches time series data set\n\n\n\n3 New Applied Tutorials:\n\nSales Analysis Tutorial\nFinance Analysis Tutorial\nDemand Forecasting Tutorial\nAnomaly Detection Tutorial\n\n\n\nFinal Deprecations:\n\nsummarize_by_time(): kind = \"period\". This was removed for consistency with pytimetk. “timestamp” is the default.\naugment_rolling(): use_independent_variables. This is replaced by augment_rolling_apply().\n\n\n\n\npytimetk 0.1.0 (2023-10-02)\n\nAbout the Initial release.\nThis release includes the following features:\n\nA workhorse plotting function called plot_timeseries() 💪\nThree (3) data wrangling functions that will simplify 90% of time series tasks 🙏\nFive (5) “augmentor” functions: These add hundreds of features to time series to help in predictive tasks 🧠\nTwo (2) time series feature summarizes: identify key aspects of your time series 🔍\nNine (9) pandas series and DatetimeIndex helpers (work more easily with these timestamp data structures) ⏲\nFour (4) date utility functions that fill in missing function gaps in pandas 🐼\nTwo (2) Visualization utilities to help you customize your visualizations and make them look MORE professional 📈\nTwo (2) Pandas helpers that help clean up and understand pandas data frames with time series 🎇\nTwelve (12) time series datasets that you can practice PyTimeTK time series analysis on 🔢\n\n\n\nThe PyTimeTK website comes with:\n\nTwo (2) Getting started tutorials\nFive (5) Guides covering common tasks\nComing Soon: Applied Tutorials in Sales, Finance, Demand Forecasting, Anomaly Detection, and more."
   },
   {
     "objectID": "reference/plot_timeseries.html",