我最近從Stata遷移到R.我不確定如何對分組和未分組的觀察結果進行計算描述性統計。分組和未分組觀察的描述性統計
這裏是我的數據:
dput(DF)
structure(list(Product_Name = c("iPhone", "iPhone", "iPhone",
"iPhone", "iPhone", "iPhone", "Nexus 6P", "Nexus 6P", "Nexus 6P",
"Nexus 6P", "Nexus 6P", "Nexus 6P"), Product_Type = c("New",
"New", "Refurbished", "New", "New", "Refurbished", "Refurbished",
"Refurbished", "Refurbished", "Refurbished", "Refurbished", "Refurbished"
), Year = c(2006, 2011, 2009, 2008, 2011, 2009, 2012, 2007, 2013,
2015, 2009, 2010), Units = c(100, 200, 300, 400, 500, 600, 700,
200, 120, 125, 345, 340)), .Names = c("Product_Name", "Product_Type",
"Year", "Units"), row.names = c(NA, 12L), class = "data.frame")
我的數據具有通過一年型銷售的產品。每個產品都可以是翻新產品或新產品。此外,如果它是在2010年之前出售的,我會將其標記爲「時間1」中出售的,否則我會將其標記爲「時間2」中出售的時間。
這裏是我的代碼如下:
DF[DF$Year<2010,"Time"]<-"1"
DF[DF$Year>=2010,"Time"]<-"2"
現在,我要生成這些時間段的描述性統計。
DF %>%
group_by(Product_Name, Product_Type,Time) %>%
dplyr::summarise(Count = n(),
Sum_Units = sum(Units,na.rm=TRUE),
Avg_Units = mean(Units,na.rm = TRUE),
Max_Units=max(Units,na.rm = TRUE))
如果我們運行上面的代碼,我們將通過Product_Name
,Product_Type
獲得描述性統計和Time
(即分組描述性統計)。但是,這不是我想要的。我需要描述性統計信息,但不考慮Product_Type
和Time
的分組。意思是,假設產品在時間1或時間2(即所有年份)內銷售並且與銷售產品的類型無關,我想要計算描述性統計量,同時保留上面的一些分組信息。
預期輸出:
dput(DFOut)
structure(list(Product_Name = c("iPhone", "Nexus 6P"), New_Units_Sum_Time1 = c(500,
NA), Refurbished_Units_Sum_Time_1 = c(900, 545), Sum_Units_Time1 = c(1400,
545), Sum_Units_Time2 = c(700, 1285), Sum_Units_Time_1_And_2 = c(2100,
1830), Avg_Units_Time1 = c(350, 272.5), Avg_Units_Time2 = c(350,
321.25), Avg_Units_Time_1_And_2 = c(350, 305), Max_Units_Time1 = c(600,
345), Max_Units_Time2 = c(500, 700), Max_Units_Time_1_And_2 = c(600,
700)), .Names = c("Product_Name", "New_Units_Sum_Time1", "Refurbished_Units_Sum_Time_1",
"Sum_Units_Time1", "Sum_Units_Time2", "Sum_Units_Time_1_And_2",
"Avg_Units_Time1", "Avg_Units_Time2", "Avg_Units_Time_1_And_2",
"Max_Units_Time1", "Max_Units_Time2", "Max_Units_Time_1_And_2"
), row.names = 1:2, class = "data.frame")
在輸出中,你會看到我有一些描述性統計:
一)根據產品的類型和它出售的時間(例如New_Units_Sum_Time1
即New
和Time1
)。請注意,在輸出中,我只顯示了New
和Time1
的組合。如果您可以指導我如何爲Refurbished
和Time
的其他組合生成描述性統計信息,那將非常棒。基於忽視產品的類型,但是不忽略)的基礎上既忽略的產品類型和時期的它被賣給(例如Sum_Units_Time_1_And_2
)出售的週期(例如Sum_Units_Time1
即Time1
)
Ç
b)中。
同上和平均值。
我該怎麼做?我會很感激任何幫助。我真的很努力。
請注意,我使用Excel手動創建了DFOut。雖然我三重檢查過,但可能會有一些手動錯誤 - 如果有問題,我會很樂意澄清它們。感謝您的時間。
sessionInfo()
R version 3.3.2 (2016-10-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)
locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252
attached base packages:
[1] grDevices datasets stats graphics grid tcltk utils methods base
other attached packages:
[1] tables_0.8 Hmisc_4.0-2 Formula_1.2-1 survival_2.40-1
[5] ResourceSelection_0.3-0 magrittr_1.5 stringr_1.1.0 bit64_0.9-5
[9] bit_1.1-12 tufterhandout_1.2.1 knitr_1.15.1 rmarkdown_1.3
[13] tufte_0.2 corrplot_0.77 purrr_0.2.2 readr_1.0.0
[17] tibble_1.2 tidyverse_1.1.1 cowplot_0.7.0 plotly_4.5.6
[21] ggplot2_2.2.1 maps_3.1.1 directlabels_2015.12.16 tidyr_0.6.1
[25] ggthemes_3.3.0 R2HTML_2.3.2 lubridate_1.6.0 xts_0.9-7
[29] zoo_1.7-14 lattice_0.20-34 corrgram_1.10 hexbin_1.27.1
[33] sm_2.2-5.4 compare_0.2-6 installr_0.18.0 psych_1.6.12
[37] reshape2_1.4.2 readstata13_0.8.5 pastecs_1.3-18 boot_1.3-18
[41] vcd_1.4-3 car_2.1-4 xlsxjars_0.6.1 rJava_0.9-8
[45] debug_1.3.1 dplyr_0.5.0 foreign_0.8-67 gmodels_2.16.2
[49] openxlsx_4.0.0 plyr_1.8.4
loaded via a namespace (and not attached):
[1] minqa_1.2.4 colorspace_1.3-2 class_7.3-14 modeltools_0.2-21 mclust_5.2.2
[6] rprojroot_1.2 htmlTable_1.9 base64enc_0.1-3 MatrixModels_0.4-1 flexmix_2.3-13
[11] mvtnorm_1.0-5 xml2_1.1.1 codetools_0.2-15 splines_3.3.2 mnormt_1.5-5
[16] robustbase_0.92-7 jsonlite_1.2 nloptr_1.0.4 pbkrtest_0.4-6 broom_0.4.1
[21] cluster_2.0.5 kernlab_0.9-25 httr_1.2.1 backports_1.0.5 assertthat_0.1
[26] Matrix_1.2-7.1 lazyeval_0.2.0 acepack_1.4.1 htmltools_0.3.5 quantreg_5.29
[31] tools_3.3.2 gtable_0.2.0 Rcpp_0.12.9 trimcluster_0.1-2 gdata_2.17.0
[36] nlme_3.1-128 iterators_1.0.8 fpc_2.1-10 lmtest_0.9-34 lme4_1.1-12
[41] rvest_0.3.2 gtools_3.5.0 dendextend_1.4.0 DEoptimR_1.0-8 MASS_7.3-45
[46] scales_0.4.1 TSP_1.1-4 hms_0.3 parallel_3.3.2 SparseM_1.74
[51] RColorBrewer_1.1-2 gridExtra_2.2.1 rpart_4.1-10 latticeExtra_0.6-28 stringi_1.1.2
[56] gclus_1.3.1 mvbutils_2.7.4.1 foreach_1.4.3 checkmate_1.8.2 seriation_1.2-1
[61] caTools_1.17.1 prabclus_2.2-6 bitops_1.0-6 evaluate_0.10 htmlwidgets_0.8
[66] R6_2.2.0 gplots_3.0.1 DBI_0.5-1 haven_1.0.0 whisker_0.3-2
[71] mgcv_1.8-16 nnet_7.3-12 modelr_0.1.0 KernSmooth_2.23-15 viridis_0.3.4
[76] readxl_0.1.1 data.table_1.10.0 forcats_0.2.0 digest_0.6.12 diptest_0.75-7
[81] stats4_3.3.2 munsell_0.4.3 registry_0.3 viridisLite_0.1.3 quadprog_1.5-5
因此,基本上你每次需要不同的分組變量......?一系列'group_by%>%...%>%ungroup()....%>%... group_by..'? – Sotos
@Sotos - 再次感謝您的幫助。是的,我要補充的唯一解釋是有多個層次:a)根據產品名稱,時間和類型分組的描述性統計b)根據產品名稱,時間分組c)根據產品名稱分組。這有幫助嗎? – watchtower
嗯,是的。這正是我所理解的:) – Sotos