Ajout de la vignette ORR #67
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,248 @@ | ||
| #' @importFrom dplyr select | ||
| #' @importFrom GenBinomApps clopper.pearson.ci | ||
| #' @importFrom flextable as_flextable surround delete_rows footnote as_paragraph | ||
| #' @importFrom officer fp_border | ||
| #' @importFrom cli cli_abort | ||
|
|
||
| ORR_table = function(id = recist$SUBJID, global_response = recist$RCRESP, date = recist$RCDT, confirmed = FALSE, show_CBR = FALSE){ | ||
| `%notin%` <- Negate(`%in%`) | ||
|
There was a problem hiding this comment. Il y a vraiment besoin d'un opérateur en plus ? if(x %notin% b) ...
if(! x %in% b) ... |
||
| if(length(id) ==0){ | ||
| cli_abort("id must be defined") | ||
| } | ||
| if(length(date) ==0){ | ||
| cli_abort("date must be defined") | ||
| } | ||
| if(length(global_response) ==0){ | ||
| cli_abort("global_response must be defined") | ||
| } | ||
| if(length(id) != length(global_response) | length(id) != length(date) | length(date) != length(global_response)){ | ||
| cli_abort("id, global_reponse and date should have the same length") | ||
| } | ||
| if(length(confirmed) > 1){ | ||
| cli_abort("confirmed shoulb be TRUE or FALSE (default = FALSE)") | ||
| } | ||
| if(length(show_CBR) > 1){ | ||
| cli_abort("show_CBR shoulb be TRUE or FALSE (default = FALSE)") | ||
| } | ||
| if(confirmed %notin% c(TRUE,FALSE,NA)){ | ||
|
There was a problem hiding this comment. tu peux utiliser les fonctions du fichier assertion.R |
||
| cli_abort("confirmed shoulb be TRUE or FALSE (default = FALSE)") | ||
| } | ||
| if(show_CBR %notin% c(TRUE,FALSE,NA)){ | ||
| cli_abort("show_CBR shoulb be TRUE or FALSE (default = FALSE)") | ||
| } | ||
| if(is.Date(date)!= TRUE){ | ||
| cli_abort("date shoul be in as.Date format") | ||
| } | ||
|
|
||
| data = data.frame(subjid = id, rcresp = global_response, rcdt = date) | ||
|
|
||
| #Si un patient a une seule visite recist (la premiere) puis aucune autre, on le modifie en Not evaluable | ||
| data = data %>% | ||
| distinct(subjid,rcresp,rcdt) %>% | ||
| arrange(as.numeric(subjid), rcdt) %>% | ||
|
There was a problem hiding this comment. cette ligne plantera dans toutes les études où SUBJID n'est pas numérique |
||
| mutate(n = n(), .by = subjid) %>% | ||
| mutate(rcresp = ifelse(n==1, "Not evaluable",as.character(rcresp))) | ||
|
|
||
| data$rcresp <- factor(data$rcresp, levels = c("Complete response", "Partial response", "Stable disease", "Progressive disease", "Not evaluable")) | ||
|
|
||
|
|
||
| recist_2 = data %>% | ||
| mutate(rcresp_num=as.numeric(as.factor(rcresp))) %>% | ||
|
There was a problem hiding this comment. Attention, tu fais des hypothèses fortes sur le contenu de |
||
| filter(!is.na(rcresp_num) & !is.na(rcdt) & !is.na(subjid)) %>% | ||
| mutate(previous_rcresp_num=lag(rcresp_num), | ||
| previous_date=lag(rcdt), | ||
| delta_date=as.numeric(rcdt - previous_date), | ||
| delta_date= ifelse(is.na(delta_date),0,delta_date), | ||
|
There was a problem hiding this comment. essaie d'espacer un peu ton code : espaces avant et après les |
||
| delta_date_before_PD_or_end = cumsum(delta_date), | ||
| delta_date_before_PD_or_end = ifelse(rcresp=="Progressive disease" ,0,delta_date_before_PD_or_end), | ||
| delta_date_before_PD_or_end= replace_na(delta_date_before_PD_or_end,0), | ||
| duree_suivi_max = max(delta_date_before_PD_or_end), | ||
| bestresponse_withinprotocole=ifelse(previous_rcresp_num==rcresp_num, 1, 0 ), | ||
| .by = subjid) | ||
|
|
||
| if (is.na(confirmed) | confirmed == FALSE){ | ||
| final_best_response=recist_2 %>% | ||
| mutate(bestresponse=min(rcresp_num), | ||
| .by = subjid) %>% | ||
| filter(bestresponse==rcresp_num) %>% | ||
| slice_head(by=subjid) %>% | ||
|
There was a problem hiding this comment. je crois que tu peux remplacer ces 4 lignes par |
||
| mutate(Overall_ORR= ifelse(rcresp=="Complete response" | rcresp=="Partial response",1,0), | ||
| CBR = ifelse(duree_suivi_max >= 152 | rcresp=="Complete response" | rcresp=="Partial response",1,0)) | ||
|
There was a problem hiding this comment. Pourquoi |
||
| } | ||
| else if(confirmed == TRUE){ | ||
| recist_2 <- recist_2 %>% | ||
| mutate( | ||
| meilleur_reponse = case_when( | ||
| rcresp_num == 1 & previous_rcresp_num == 1 & delta_date >= 28 ~ 1, | ||
| rcresp_num == 1 & previous_rcresp_num == 1 & delta_date < 28 ~ 3, | ||
| rcresp_num == 1 & previous_rcresp_num == 2 & delta_date >= 28 ~ 2, | ||
| rcresp_num == 1 & previous_rcresp_num == 2 & delta_date < 28 ~ 3, | ||
| rcresp_num == 1 & previous_rcresp_num == 3 ~ 3, | ||
| rcresp_num == 1 & previous_rcresp_num == 4 ~ 4, | ||
| rcresp_num == 1 & previous_rcresp_num == 5 ~ 5, | ||
|
|
||
| rcresp_num == 2 & previous_rcresp_num <= 2 & delta_date >= 28 ~ 2, | ||
| rcresp_num == 2 & previous_rcresp_num <= 2 & delta_date < 28 ~ 3, | ||
| rcresp_num == 2 & previous_rcresp_num == 3 ~ 3, | ||
| rcresp_num == 2 & previous_rcresp_num == 4 ~ 4, | ||
| rcresp_num == 2 & previous_rcresp_num == 5 ~ 5, | ||
|
|
||
| is.na(previous_rcresp_num) & rcresp_num == 4 ~ 4, | ||
|
|
||
| TRUE ~ rcresp_num | ||
| ) | ||
| ) | ||
|
|
||
| final_best_response=recist_2 %>% | ||
| mutate(bestresponse=min(meilleur_reponse), | ||
| .by=subjid | ||
| ) %>% | ||
| filter(bestresponse==meilleur_reponse) %>% | ||
| slice_head(by=subjid) %>% | ||
| mutate(Overall_ORR= ifelse(rcresp=="Complete response" | rcresp=="Partial response",1,0), | ||
| CBR = ifelse(duree_suivi_max >= 152 | rcresp=="Complete response" | rcresp=="Partial response",1,0)) | ||
| } | ||
|
|
||
| total <- length(unique(final_best_response$subjid)) | ||
| CR <- length(final_best_response$rcresp[final_best_response$rcresp == "Complete response"]) | ||
| PR <- length(final_best_response$rcresp[final_best_response$rcresp == "Partial response"]) | ||
| SD <- length(final_best_response$rcresp[final_best_response$rcresp == "Stable disease"]) | ||
| PD <- length(final_best_response$rcresp[final_best_response$rcresp == "Progressive disease"]) | ||
| NE <- length(final_best_response$rcresp[final_best_response$rcresp == "Not evaluable"]) | ||
| Overall_ORR <- CR + PR | ||
| CBR <- length(final_best_response$CBR[final_best_response$CBR==1]) | ||
|
|
||
| CR_CP <- clopper.pearson.ci(CR,total,CI="two.sided", alpha = 0.05) | ||
| PR_CP <- clopper.pearson.ci(PR,total,CI="two.sided", alpha = 0.05) | ||
| SD_CP <- clopper.pearson.ci(SD,total,CI="two.sided", alpha = 0.05) | ||
| PD_CP <- clopper.pearson.ci(PD,total,CI="two.sided", alpha = 0.05) | ||
| NE_CP <- clopper.pearson.ci(NE,total,CI="two.sided", alpha = 0.05) | ||
| Overall_ORR_CP <- clopper.pearson.ci(Overall_ORR,total,CI="two.sided", alpha = 0.05) | ||
| CBR_CP <- clopper.pearson.ci(CBR,total,CI="two.sided", alpha = 0.05) | ||
|
|
||
| CR_CP_IC <- paste0("[",round(CR_CP$Lower.limit*100,1),";",round(CR_CP$Upper.limit*100,1),"]") | ||
| PR_CP_IC <- paste0("[",round(PR_CP$Lower.limit*100,1),";",round(PR_CP$Upper.limit*100,1),"]") | ||
| SD_CP_IC <- paste0("[",round(SD_CP$Lower.limit*100,1),";",round(SD_CP$Upper.limit*100,1),"]") | ||
| PD_CP_IC <- paste0("[",round(PD_CP$Lower.limit*100,1),";",round(PD_CP$Upper.limit*100,1),"]") | ||
| NE_CP_IC <- paste0("[",round(NE_CP$Lower.limit*100,1),";",round(NE_CP$Upper.limit*100,1),"]") | ||
| Overall_ORR_CP_IC <- paste0("[",round(Overall_ORR_CP$Lower.limit*100,1),";",round(Overall_ORR_CP$Upper.limit*100,1),"]") | ||
| CBR_CP_IC <- paste0("[",round(CBR_CP$Lower.limit*100,1),";",round(CBR_CP$Upper.limit*100,1),"]") | ||
|
There was a problem hiding this comment. Attention, beaucoup de baseR et beaucoup de répétitions (DRY), donc risque d'erreur et mauvaise maintenabilité.
|
||
|
|
||
| if (is.na(show_CBR) | show_CBR == FALSE){ | ||
| IC_95 <- c(Overall_ORR_CP_IC, | ||
| CR_CP_IC, | ||
| PR_CP_IC, | ||
| SD_CP_IC, | ||
| PD_CP_IC, | ||
| NE_CP_IC) | ||
|
There was a problem hiding this comment. Bon alors là je vais dire le contraire, c'est un peu trop aéré 😂 |
||
|
|
||
| Name <- c("Overall ORR", | ||
| "Complete response (CR)", | ||
| "Partial response (PR)", | ||
| "Stable disease (SD)", | ||
| "Progressive disease (PD)", | ||
| "Not evaluable (NE)") | ||
|
|
||
| N <- c(Overall_ORR, | ||
| CR, | ||
| PR, | ||
| SD, | ||
| PD, | ||
| NE) | ||
|
|
||
| Percentage <- c(round(Overall_ORR/total*100,1), | ||
| round(CR/total*100,1), | ||
| round(PR/total*100,1), | ||
| round(SD/total*100,1), | ||
| round(PD/total*100,1), | ||
| round(NE/total*100,1)) | ||
| } | ||
| else if(show_CBR == TRUE){IC_95 <- c(Overall_ORR_CP_IC, | ||
| CR_CP_IC, | ||
| PR_CP_IC, | ||
| SD_CP_IC, | ||
| PD_CP_IC, | ||
| NE_CP_IC, | ||
| CBR_CP_IC) | ||
|
|
||
| Name <- c("Overall ORR", | ||
| "Complete response (CR)", | ||
| "Partial response (PR)", | ||
| "Stable disease (SD)", | ||
| "Progressive disease (PD)", | ||
| "Not evaluable (NE)", | ||
| "Clinical Benefit Rate (CBR)") | ||
|
|
||
| N <- c(Overall_ORR, | ||
| CR, | ||
| PR, | ||
| SD, | ||
| PD, | ||
| NE, | ||
| CBR) | ||
|
|
||
| Percentage <- c(round(Overall_ORR/total*100,1), | ||
| round(CR/total*100,1), | ||
| round(PR/total*100,1), | ||
| round(SD/total*100,1), | ||
| round(PD/total*100,1), | ||
| round(NE/total*100,1), | ||
| round(CBR/total*100,1)) | ||
| } | ||
|
|
||
| Best_Response_during_treatment <- as.data.frame(cbind(Name,N,Percentage,IC_95)) | ||
| if (is.na(confirmed) | confirmed == FALSE){ | ||
| nom_col = c("Unconfirmed Best Response during treatment",paste0("N=",total),"%","IC 95%") | ||
| } | ||
| else if(confirmed == TRUE){ | ||
| nom_col = c("Confirmed Best Response during treatment",paste0("N=",total),"%","IC 95%") | ||
| } | ||
|
|
||
| colnames(Best_Response_during_treatment) <- nom_col | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
|
There was a problem hiding this comment. Il faut toujours séparer le fond et la forme. |
||
| as_flextable(show_coltype = F, include.row_percent = FALSE, include.column_percent = FALSE, include.table_percent = FALSE) %>% | ||
|
There was a problem hiding this comment.
|
||
| delete_rows(i=1, part = "footer") %>% | ||
| bold(bold = TRUE, part = "header") %>% | ||
| surround(i = c(1,6), j = 1:4, border.bottom = fp_border(color = "black", style = "solid", width = 1), part = "body") | ||
|
|
||
| if (is.na(confirmed) | confirmed == FALSE){ | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
| footnote( i = 1, j = c(4), | ||
| value = as_paragraph( | ||
| c("Clopper-Pearson (Exact) method was used for confidence interval")), | ||
| ref_symbols =c("*"), part = "header") %>% | ||
| valign(valign = "bottom", part = "header") | ||
| } | ||
| else if(confirmed == TRUE){ | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
| footnote( i = 1, j = c(4,1), | ||
| value = as_paragraph( | ||
| c("Clopper-Pearson (Exact) method was used for confidence interval", | ||
| "For CR & PR confirmation of response had to be be demonstrated with an assessment 4 weeks or later from the initial response for response. *As stated in the protocol: «For equivocal findings of progression (e.g., very small and uncertain new lesions; cystic changes or necrosis in existing lesions), treatment may continue until the next scheduled assessment». Therefore, some patients had a response after a progressive disease response, in that case best response was examine before and after PD. Scans conducted after initiating new anti-cancer therapy were not included in the ORR analyses")), | ||
| ref_symbols =c("*","1"), part = "header") %>% | ||
| valign(valign = "bottom", part = "header") | ||
| } | ||
|
|
||
| if (is.na(show_CBR) | show_CBR == FALSE){ | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
| bold(i = c(1), j = 1, bold = TRUE, part = "body") | ||
| } | ||
| else if(show_CBR == TRUE & confirmed == TRUE){ | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
| footnote( i = c(7), j = 1, | ||
| value = as_paragraph( | ||
| c("CBR was defined as the presence of at least a partial response (PR), complete response (CR), or stable disease (SD) lasting at least six months (using a window of +/-1 month for the RECIST date).")), | ||
| ref_symbols = c("2"), part = "body") %>% | ||
| valign(valign = "bottom", part = "header") %>% | ||
| bold(i = c(1,7), j = 1, bold = TRUE, part = "body") | ||
| } | ||
| else if(show_CBR == TRUE & (is.na(confirmed) | confirmed == FALSE)){ | ||
| Best_Response_during_treatment = Best_Response_during_treatment %>% | ||
| footnote( i = c(7), j = 1, | ||
| value = as_paragraph( | ||
| c("CBR was defined as the presence of at least a partial response (PR), complete response (CR), or stable disease (SD) lasting at least six months (using a window of +/-1 month for the RECIST date).")), | ||
| ref_symbols = c("1"), part = "body") %>% | ||
| valign(valign = "bottom", part = "header") %>% | ||
| bold(i = c(1,7), j = 1, bold = TRUE, part = "body") | ||
| } | ||
| Best_Response_during_treatment | ||
| } | ||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Dans grstat, on utilise des interfaces
tidyverse, pas base R comme iciça veut dire que le premier argument est quasi toujours une dataframe
donc il vaut mieux avoir dans tes arguments d'abord
data, puis un moyen d'identifier les colonnes, par exemplesubjid="SUBJID", resp="RCRESP".Attention, tes valeurs par défaut attendent qu'un objet
recistexiste, ce qui fera planter la fonction dans quasi tous les cas de figures réels.