Message Boards

Answer (Unmark)

GROUPS:

Staff Picks Biological Sciences Data Science Medical Sciences Graphics and Visualization Wolfram Language Modeling Statistics and Probability Machine Learning Events & Media

Tobias Canavesi

[NB] Argentina: COVID-19 Data Analysis

Tobias Canavesi, Physics Institute of La Plata

Posted 2 years ago

5845 Views

3 Replies

4 Total Likes

Follow this post

MODERATOR NOTE: coronavirus resources & updates: https://wolfr.am/coronavirus

enter image description here

See visualizations that I did on number of Covid-19 cases in Argentina. I also did some tests using Predict function with different methods.

Data on Covid19 infections in Argentina

In[]:=

(*UpdatedApril3,2020*)casosarg={{5,1},{6,2},{7,9},{8,12},{9,17},{10,19},{11,21},{12,31},{13,34},{14,45},{15,56},{16,65},{17,79},{18,97},{19,129},{20,158},{21,225},{22,266},{23,301},{24,387},{25,502},{26,589},{27,690},{28,745},{29,820},{30,966},{31,1054},{32,1133},{33,1265},{34,1353},{35,1451}};(*Consideringadataerrorofabout5percent*)errorescasoarg=Table[casosarg[[i,2]]*0.05,{i,1,Length[casosarg]}];corona=ListPlot[casosarg,PlotStyleRed]

Out[]=

In[]:=

datamatrix=Import["https://github.com/tobiascanavesi/covidarg/blob/master/datoscovid.txt?raw=true","Table",CharacterEncoding"UTF8"]//MatrixForm;

In[]:=

Analysis of different methods of the predict function

In[]:=

(*Comparisonsofpredictivefunctionsusing26cases*)

In[]:=

casosargtraining=Table[i+4->casosarg[[i,2]],{i,Length[casosarg[[1;;26]]]}];

In[]:=

p0=Predict[casosargtraining,Method"NeuralNetwork"]

Out[]=

PredictorFunction

Input type: Numerical

Method: NeuralNetwork



In[]:=

Show[Plot[{p0[x],p0[x]+StandardDeviation[p0[x,"Distribution"]],p0[x]-StandardDeviation[p0[x,"Distribution"]]},{x,27,35},PlotStyle{Blue,Gray,Gray},Filling{2{3}},ExclusionsFalse,PerformanceGoal"Speed",PlotLegends{"Prediction Neural Network","Confidence Interval"}],ListPlot[List@@@casosarg[[27;;31]],PlotStyleRed,PlotLegends{"Data"}]]

Out[]=

	Prediction Neural Network
	Confidence Interval

Data

In[]:=

p1=Predict[casosargtraining,Method->"GradientBoostedTrees"]

Out[]=

PredictorFunction

Input type: Numerical

Method: GradientBoostedTrees



In[]:=

Show[Plot[{p1[x],p1[x]+StandardDeviation[p1[x,"Distribution"]],p1[x]-StandardDeviation[p1[x,"Distribution"]]},{x,27,35},PlotStyle{Blue,Gray,Gray},Filling{2{3}},ExclusionsFalse,PerformanceGoal"Speed",PlotLegends{"Prediction Gradient Boosted Trees","Confidence Interval"}],ListPlot[List@@@casosarg[[27;;31]],PlotStyleRed,PlotLegends{"Data"}]]

Out[]=

	Prediction Gradient Boosted Trees
	Confidence Interval

Data

In[]:=

p2=Predict[casosargtraining,Method->"DecisionTree"]

Out[]=

PredictorFunction

Input type: Numerical

Method: DecisionTree



In[]:=

Show[Plot[{p2[x],p2[x]+StandardDeviation[p2[x,"Distribution"]],p2[x]-StandardDeviation[p2[x,"Distribution"]]},{x,27,35},PlotStyle{Blue,Gray,Gray},Filling{2{3}},ExclusionsFalse,PerformanceGoal"Speed",PlotLegends{"Prediction Decision Tree","Confidence Interval"}],ListPlot[List@@@casosarg[[27;;31]],PlotStyleRed,PlotLegends{"Data"}]]

Out[]=

	Prediction Decision Tree
	Confidence Interval

Data

In[]:=

p3=Predict[casosargtraining,Method->"RandomForest"]

Out[]=

PredictorFunction

Input type: Numerical

Method: RandomForest



In[]:=

Show[Plot[{p3[x],p3[x]+StandardDeviation[p3[x,"Distribution"]],p3[x]-StandardDeviation[p3[x,"Distribution"]]},{x,27,35},PlotStyle{Blue,Gray,Gray},Filling{2{3}},ExclusionsFalse,PerformanceGoal"Speed",PlotLegends{"Prediction Random Forest","Confidence Interval"}],ListPlot[List@@@casosarg[[27;;31]],PlotStyleRed,PlotLegends{"Data"}]]

Out[]=

	Prediction Random Forest
	Confidence Interval

Data

In[]:=

p4=Predict[casosargtraining,Method->"GaussianProcess"]

Out[]=

PredictorFunction

Input type: Numerical

Method: GaussianProcess



In[]:=

Show[Plot[{p4[x],p4[x]+StandardDeviation[p4[x,"Distribution"]],p4[x]-StandardDeviation[p4[x,"Distribution"]]},{x,27,35},PlotStyle{Blue,Gray,Gray},Filling{2{3}},ExclusionsFalse,PerformanceGoal"Speed",PlotLegends{"Prediction Gaussian Process","Confidence Interval"}],ListPlot[List@@@casosarg[[27;;31]],PlotStyleRed,PlotLegends{"Data"}]]

Out[]=

	Prediction Gaussian Process
	Confidence Interval

Data

In[]:=

(*HereHerewecanseehoweachpredictordiddaybyday,thecloserto1thebetterthepredictorworked*)

In[]:=

differentpredicts=Table[{(Abs[p0[1+i]])/(casosarg[[i,2]]),(Abs[p1[1+i]])/(casosarg[[i,2]]),(Abs[p2[1+i]])/(casosarg[[i,2]]),(Abs[p3[1+i]])/(casosarg[[i,2]]),(Abs[p4[1+i]])/(casosarg[[i,2]])},{i,27,Length[casosarg]}]

Out[]=

{{0.711277,0.582385,0.72296,0.611265,0.71797},{0.740969,0.541778,0.672551,0.568644,0.751586},{0.734708,0.485244,0.602372,0.509307,0.745541},{0.752295,0.453684,0.563193,0.476181,0.760129},{0.760366,0.423042,0.525155,0.44402,0.76138}}

In[]:=

neuralplot=Show[ListPlot[Table[differentpredicts[[i,1]],{i,1,Length[differentpredicts]}]],Plot[Evaluate[y=1],{x,0,5},PlotStyleRed],PlotLabel"Neural Network",PlotRange->All]

Out[]=

In[]:=

Mean[Table[differentpredicts[[i,1]],{i,1,Length[differentpredicts]}]]

Out[]=

0.739923

In[]:=

gbtplot=Show[ListPlot[Table[differentpredicts[[i,2]],{i,1,Length[differentpredicts]}]],Plot[Evaluate[y=1],{x,0,5},PlotStyleRed],PlotLabel"Gradient Boosted Trees"]

Out[]=

In[]:=

Mean[Table[differentpredicts[[i,2]],{i,1,Length[differentpredicts]}]]

Out[]=

0.497227

In[]:=

pdtplot=Show[ListPlot[Table[differentpredicts[[i,3]],{i,1,Length[differentpredicts]}]],Plot[Evaluate[y=1],{x,0,5},PlotStyleRed],PlotLabel"Prediction Decision Tree"]

Out[]=

In[]:=

Mean[Table[differentpredicts[[i,3]],{i,1,Length[differentpredicts]}]]

Out[]=

0.617246

In[]:=

prfplot=Show[ListPlot[Table[differentpredicts[[i,4]],{i,1,Length[differentpredicts]}]],Plot[Evaluate[y=1],{x,0,5},PlotStyleRed],PlotLabel"Prediction Random Forest"]

Out[]=

In[]:=

Mean[Table[differentpredicts[[i,4]],{i,1,Length[differentpredicts]}]]

Out[]=

0.521884

In[]:=

pgpplot=Show[ListPlot[Table[differentpredicts[[i,5]],{i,1,Length[differentpredicts]}]],Plot[Evaluate[y=1],{x,0,5},PlotStyleRed],PlotLabel"Prediction Gaussian Process",PlotRange->All]

Out[]=

In[]:=

Mean[Table[differentpredicts[[i,5]],{i,1,Length[differentpredicts]}]]

Out[]=

0.747321

In[]:=

(*Thetwobestpredictorsare"Neural Network and Gaussian Process"*)

Figures on the progress of the Covid19 in Argentina

DATA

In[]:=

contagios2=GeoRegionValuePlot

Buenos Aires, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,4,3]],

Buenos Aires

CITY

datamatrix[[1,3,5]],

Rio Negro, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,5,3]],

Cordoba, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,6,2]],

San Luis, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,7,3]],

Santa Fe, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,8,3]],

Chaco, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,9,2]],

Tierra del Fuego, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,10,4]],

Jujuy, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,11,2]],

Salta, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,12,2]],

Entre Rios, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,13,3]],

Santa Cruz, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,14,3]],

Tucuman, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,15,2]],

Corrientes, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,16,2]],

Neuquen, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,17,2]],

Santiago del Estero, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,18,4]],

Mendoza, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,19,2]],

Misiones, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,20,2]],

La Pampa, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,21,3]],

San Juan, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,22,3]],

La Rioja, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,23,3]];

In[]:=

recuperados=GeoBubbleChart

Buenos Aires, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,4,4]],

Buenos Aires

CITY

datamatrix[[1,3,6]],

Rio Negro, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,5,4]],

Cordoba, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,6,3]],

San Luis, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,7,4]],

Santa Fe, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,8,4]],

Chaco, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,9,3]],

Tierra del Fuego, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,10,5]],

Jujuy, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,11,3]],

Salta, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,12,3]],

Entre Rios, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,13,4]],

Santa Cruz, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,14,4]],

Tucuman, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,15,3]],

Corrientes, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,16,3]],

Neuquen, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,17,3]],

Santiago del Estero, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,18,5]],

Mendoza, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,19,3]],

Misiones, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,20,3]],

La Pampa, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,21,4]],

San Juan, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,22,4]],

La Rioja, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,23,4]],ChartElementFunction"TriangleWaveBubble",BubbleSizes{0.1,0.25},ChartStyleHue[.5];

muertes=GeoBubbleChart

Buenos Aires, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,4,5]],

Buenos Aires

CITY

datamatrix[[1,3,7]],

Rio Negro, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,5,5]],

Cordoba, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,6,4]],

San Luis, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,7,5]],

Santa Fe, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,8,5]],

Chaco, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,9,4]],

Tierra del Fuego, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,10,6]],

Jujuy, Argentina

ADMINISTRATIVE DIVISION

datamatrix[[1,1