SciPy 1.0: fundamental algorithms for scientific computing in Python

Virtanen, Pauli; Gommers, Ralf; Oliphant, Travis E.; Haberland, Matt; Reddy, Tyler; Cournapeau, David; Burovski, Evgeni; Peterson, Pearu; Weckesser, Warren; Bright, Jonathan; Walt, Stéfan J. van der; Brett, Matthew; Wilson, Joshua; Millman, K. Jarrod; Mayorov, Nikolay; Nelson, Andrew; Jones, Eric D.; Kern, Robert; Larson, Eric B.; Carey, CJ; Polat, İlhan; Feng, Yu; Moore, Eric W.; Vanderplas, Jake; Laxalde, Denis; Perktold, Josef; Cimrman, Robert; Henriksen, Ian; Quintero, E. A.; Harris, C. R.; Archibald, Anne M.; Ribeiro, Antônio H.; Pedregosa, Fabián; Mulbregt, Paul van; Vijaykumar, A.; Bardelli, Alessandro Pietro; Rothberg, Alex; Hilboll, Andreas; Kloeckner, Andreas; Scopatz, Anthony; Lee, Antony; Rokem, Ariel; Woods, C. Nathan; Fulton, Chad; Masson, Charles; Häggström, Christian; Fitzgerald, Clark; Nicholson, David; Hagen, David; Ṗasechnik, Dmitrii V.; Olivetti, Emanuele; Martin, Éric; Wieser, Eric; Silva, Fabrice; Lenders, Felix; Wilhelm, Florian; Young, George S.; Price, Gavin A.; Ingold, Gert‐Ludwig; Allen, Gregory E.; Lee, Gregory R.; Audren, Hervé; Probst, Irvin; Dietrich, J. P.; Silterra, Jacob; Webber, James T.; Slavič, Janko; Nothman, Joel; Buchner, Johannes; Kulick, Jeffrey H.; Schönberger, Johannes L.; Cardoso, José Vinícius de Miranda; Reimer, Joscha; Harrington, Joseph; Rodríguez, Juan Luis; Nunez-Iglesias, Juan; Kuczynski, Justin; Tritz, K.; Thoma, Martin; Newville, Matthew; Kümmerer, Matthias; Bolingbroke, Maximilian; Tartre, Michael; Pak, Mikhail; Smith, Nathaniel J.; Nowaczyk, Nikolai; Shebanov, Nikolay; Pavlyk, Oleksandr; Brodtkorb, Per A.; Lee, Perry; McGibbon, Robert T.; Feldbauer, Roman; Lewis, Sam M.; Tygier, Sam; Sievert, Scott; Vigna, Sebastiano; Peterson, Stefan; More, Surhud; Pudlik, Tadeusz; Oshima, Takuya; Pingel, Thomas J.; Robitaille, T. P.; Spura, Thomas; Jones, Thouis R.; Cera, Tim; Leslie, Tim; Zito, Tiziano; Krauss, Tom; Upadhyay, Utkarsh; Halchenko, Yaroslav; Vázquez-Baeza, Yoshiki

doi:10.1038/s41592-019-0686-2

Cited by 25,257 publications

(12,960 citation statements)

References 76 publications

Supporting

Mentioning

12,792

Contrasting

Unclassified

147

Order By: Relevance

“…Then, in order to avoid the problem of the presence of local minima for r eff in the least-square fitting algorithm, we generate models on a grid of radii of 0.01 μm between 0.1 and 8 μm, and for each one, we retrieve the optimal value of using the Trust Region Reflective algorithm implemented in the scipy.optimize.curve_fit() Python function (Virtanen et al, 2019). Thus, we obtain the best fit for each particle size, which we compare in a second time, by computing a 2 from the data and the model, as defined in equation (9) (Bevington & Robinson, 1992).…”

Section: Water Ice Particle Size Retrieving Methodsmentioning

confidence: 99%

Martian Water Ice Clouds During the 2018 Global Dust Storm as Observed by the ACS‐MIR Channel Onboard the Trace Gas Orbiter

et al. 2020

View full text Add to dashboard Cite

The Atmospheric Chemistry Suite (ACS) instrument onboard the ExoMars Trace Gas Orbiter (TGO) European Space Agency‐Roscosmos mission began science operations in March 2018. ACS Mid‐InfraRed (MIR) channel notably provides solar occultation observations of the Martian atmosphere in the 2.3‐ to 4.2‐ normalμ m spectral range. Here, we use these observations to characterize water ice clouds before and during the MY 34 Global Dust Storm (GDS). We developed a method to detect water ice clouds with mean particle size ≤ 2 normalμ m and applied it to observations gathered between Ls=165∘ and Ls=243∘. We observe a shift in water ice cloud maximum altitudes from about 60 km before the GDS to above 90 km during the storm. These very high altitude, small‐sized ( reff≤0.3.3emnormalμ m) water ice clouds are more frequent during MY 34 compared to non‐GDS years at the same season. Particle size frequently decreases with altitude, both locally within a given profile and globally in the whole data set. We observe that the maximum altitude at which a given size is observed can increase during the GDS by several tens of kilometers for certain sizes. We notably notice some large water ice particles ( reff≥1.5.3emnormalμ m) at surprisingly high altitudes during the GDS (50–70 km). These results suggest that GDS can significantly impact the formation and properties of high‐altitude water ice clouds as compared to the usual perihelion dust activity.

show abstract

Section: Water Ice Particle Size Retrieving Methodsmentioning

confidence: 99%

Martian Water Ice Clouds During the 2018 Global Dust Storm as Observed by the ACS‐MIR Channel Onboard the Trace Gas Orbiter

et al. 2020

View full text Add to dashboard Cite

show abstract

“…, for day t after behavior change and constant a and exponent factor b. To fit the data points we use the optimize.curve_f it implementation of the scipy.optimize library (13).…”

Section: B Limitationsmentioning

confidence: 99%

Correcting under-reported COVID-19 case numbers: estimating the true scale of the pandemic

Jagodnik

Ray

Giorgi

et al. 2020

Preprint

100

View full text Add to dashboard Cite

The COVID-19 virus has spread worldwide in a matter of a few months, while healthcare systems struggle to monitor and report current cases. Testing results have struggled with the relative capabilities, testing policies and preparedness of each affected country, making their comparison a non-trivial task. Since severe cases, which more likely lead to fatal outcomes, are detected at a higher rate than mild cases, the reported virus mortality is likely inflated in most countries. Lockdowns and changes in human behavior modulate the underlying growth rate of the virus. Under-sampling of infection cases may lead to the under-estimation of total cases, resulting in systematic mortality estimation biases. For healthcare systems worldwide it is important to know the expected number of cases that will need treatment. In this manuscript, we identify a generalizable growth rate decay reflecting behavioral change. We propose a method to correct the reported COVID-19 cases and death numbers by using a benchmark country (South Korea) with nearoptimal testing coverage, with considerations on population demographics. We extrapolate expected deaths and hospitalizations with respect to observations in countries that passed the exponential growth curve. By applying our correction, we predict that the number of cases is highly under-reported in most countries and a significant burden on worldwide hospital capacity.

show abstract

“…With these ranges set, we approximate the optimizer via Gaussian maximum likelihood estimation with differential evolution minimization [18]. The final result is then further optimized within our provided bounds by using the L-BFGS-B method from SciPy [19]. The quality improvements of the fit after this final local optimization step are minor, indicating that differential evolution of the maximum likelihood estimator already achieves accurate parameter discovery.…”

Section: B2 Parameter Estimationmentioning

confidence: 99%

“…The quality improvements of the fit after this final local optimization step are minor, indicating that differential evolution of the maximum likelihood estimator already achieves accurate parameter discovery. We numerically solve the differential equation system with the integrate.odeint function from SciPy that acts as a wrapper for lsoda [19,20].…”

Section: B2 Parameter Estimationmentioning

confidence: 99%

Mathematical modeling of COVID-19 containment strategies with considerations for limited medical resources

Eastman

Meaney

Przedborski

et al. 2020

Preprint

View full text Add to dashboard Cite

The outbreak of SARS-CoV-2 in China has spread around the world, infecting millions and causing governments to implement strict policies to counteract the spread of the disease. One of the most effective strategies in reducing the severity of the pandemic is social distancing, where members of the population systematically reduce their interactions with others to limit the transmission rate of the virus. However, the implementation of social distancing can be difficult and costly, making it imperative that both policy makers and the citizenry understand the potential benefits if done correctly and the risks if not. In this work, a mathematical model is developed to study the effects of social distancing on the spread of the SARS-CoV-2 virus in Canada. The model is based upon a standard epidemiological SEIRD model that has been stratified to directly incorporate the proportion of individuals who are following social distancing protocols. The model parameters characterizing the disease are estimated from current epidemiological data on COVID-19 using machine learning techniques. The results of the model show that social distancing policies in Canada have already saved thousands of lives and that the prolonged adherence to social distancing guidelines could save thousands more. Importantly, our model indicates that social distancing can significantly delay the onset of infection peaks, allowing more time for the production of a vaccine or additional medical resources. Furthermore, our results stress the importance of easing social distancing restrictions gradually, rather than all at once, in order to prevent a second wave of infections. Model results are compared to the current capacity of the Canadian healthcare system by examining the current and future number of ventilators available for use, emphasizing the need for the increased production of additional medical resources.

show abstract

SciPy 1.0: fundamental algorithms for scientific computing in Python

Cited by 25,257 publications

References 76 publications

Martian Water Ice Clouds During the 2018 Global Dust Storm as Observed by the ACS‐MIR Channel Onboard the Trace Gas Orbiter

Martian Water Ice Clouds During the 2018 Global Dust Storm as Observed by the ACS‐MIR Channel Onboard the Trace Gas Orbiter

Correcting under-reported COVID-19 case numbers: estimating the true scale of the pandemic

Mathematical modeling of COVID-19 containment strategies with considerations for limited medical resources

Contact Info

Product

Resources

About