יש ליצור מחדש את התמונה graph הזו באמצעות גרפיקה וקטורית כדוגמת קובץ SVG. לפעולה זו מספר יתרונות, כפי שניתן לקרוא (באנגלית) בדף Commons:Media for cleanup. אם ברשותכם קובץ SVG, אנא העלו אותו. לאחר מכן, החליפו תבנית זו בתבנית
תיאור2017- Donald Trump veracity - composite graph.png
English: Chart of Donald Trump's “False or misleading claims” (The Washington Post) and “False claims” by Daniel Dale (Toronto Star, later CNN)
Graph with background orange bars, "False or misleading claims", is from monthly data in the The Washington Post fact checker, and is directly represented.
Graph with foreground violet bars, "False claims", is in time segments:
1. Jan 2017–May 2019: from fact checker Daniel Dale's weekly chart published in Toronto Star. Color gradients (violet changing to orange) show the full range of possible "False Statements" resulting when the Toronto Star's weekly data was converted to monthly data. See Technical details below.
2. June 2019: No data available since Daniel Dale was between jobs.
3. July 2019— : from fact checker Daniel Dale's daily chart published on CNN. I grouped the days' values by month—of course requiring no calculation, error bars, or color gradients.
Orange values and violet values aren't "stacked" one atop the other. The violet bars are superimposed in front of the orange bars to save space.
Beginning with Version 5 (2020-02-26), Toronto Star weekly data for weeks that are split across two months is represented using COLOR GRADIENTS (gradually changing color) in the middle portion of each column of the chart through May 2019. In Version 5, the bottom of the gradient area of the column assumes NO data from such "split" week(s) belongs to that month, and the top of the gradient area assumes ALL data from such "split" week(s) belongs to that month. This new approach graphically shows the full range of possible False Statements for such months, eliminates "error bar" estimation, and avoids algebraic calculations needed to determine the center point for error bars.
In Versions 1 through 4: "Interpolation" example (used in Versions 1 through 4 for converting the Toronto Star'sweekly figures into monthly figures) In a representative example, the February 2, 2018 week in the Toronto Star indicated 36 false statements. That 7-day week includes 5 days in January and 2 days in February. Thus, that week's contribution to January is simply 36*(5/7)=25.7, and its contribution to February is 36*(2/7)=10.3. Of course, weeks entirely within a single month contribute 100% to that month.
Click at right to show/hide data for Toronto Star / CNN portions (below, inside, and above Gradient regions in Version 5)
Each three of the following data items are "stacked" (violet, below gradient, below orange) to form a full-height column.
(In the visible chart, the violet columns appear "in front of" the orange columns are not "stacked".)
MonthMINGRADIENTWashPost
Jan '17 27 15 13
Feb '17 57 32 64
Mar '17 57 17 74
Apr '17 47 8 110
May '17 32 29 44
Jun '17 81 20 48
Jul '17 117 18 92
Aug '17 62 31 50
Sep '17 85 47 39
Oct '17 113 66 66
Nov '17 43 62 115
Dec '17 105 39 71
Jan '18 81 33 175
Feb '18 92 60 39
Mar '18 147 69 45
Apr '18 90 61 138
May '18 68 61 120
Jun '18 272 76 186
Jul '18 186 174 93
Aug '18 231 132 116
Sep '18 269 129 201
Oct '18 484 369 352
Nov '18 283 240 344
Dec '18 110 78 82
Jan '19 123 164 450
Feb '19 121 197 167
Mar '19 196 217 140
Apr '19 130 136 421
May '19 234 31 311
Jun '19 674
Jul '19 242 0 302
Aug '19 247 0 280
Sep '19 280 0 359
Oct '19 393 0 763
Nov '19 244 0 656
Dec '19 209 0 475
רישיון
אני, בעל זכויות היוצרים על עבודה זו, מפרסם בזאת את העבודה תחת הרישיון הבא:
ייחוס – יש לתת ייחוס הולם, לתת קישור לרישיון, ולציין אם נעשו שינויים. אפשר לעשות את זה בכל צורה סבירה, אבל לא בשום צורה שמשתמע ממנה שמעניק הרישיון תומך בך או בשימוש שלך.
שיתוף זהה – אם תיצרו רמיקס, תשנו, או תבנו על החומר, חובה עליכם להפיץ את התרומות שלך לפי תנאי רישיון זהה או תואם למקור.
Version 11: Updating graph to include exact value for August 2020 that no longer involves extrapolation for the last four days of the month. Washington Post added a few more days to their database.
Version 10: Update to include Washington Post data for July and August 2020. Had to expand vertical scale to accommodate new high in data value (for August 2020).