200905490 (1) 九、發明說明 【發明所屬之技術領域】 本發明關係於將數値資料圖表化的方法，更明確地 說’本發明係有關於用以指出資料中，特別是財務報表資 料中’如總分類帳中之趨勢、異常、或重大變動的一種方 法。 【先前技術】 在現行的關聯性資料庫裡，資料都是以行和列的方式 來管理’只有欄位與欄位、或是資料格和資料格之間的關 係，而欠缺了前後關係與順序性，分析所需的資料整合， 完全必須靠人腦去做判斷。倘若要做與去年或是上個月的 比較，就要寫S Q L (結構化查詢語言）才能辦到，不只麻 煩、也常可能延誤了決策的時效性。 因此，市場上出現了有OLAP(線上分析處理）分析工 具，其能夠預先依照維度，來有效地分類資料，更重要的 是，它能夠建立所謂的時間維度，也就是以時間來分類資 料。時間維度的切割，可以依照資料屬性而有不同，例如 出貨量以天計算，而比較高階的報表，像是銷售趨勢看月 份、而財報則以季爲單位。 因此，在台灣專利公開第20063276 8號案中’即揭示 使用OLAP來進行含有財務資料之資料庫的法務調查’以 使資料庫使用者能輕易地可以獲得特定期間的資料量値、 然後去進行相互的比較或者是掌握趨勢的變化’藉以達到 -4- (2) (2)200905490 快速的最佳分析效果。 該案仍沿襲現今多數的管理者般都是使用EXCEL或 類似之作圖工具來作圖形表現，然而，excel所作出之 圖形表示方法，仍有著傳統統計圖表唬弄閱圖者的情形° 例如，如第7a圖之七至十二月的銷售業績，該圖的曲線 先是很快爬上最高點，然後就一直往下掉，因此，作圖者 會想要以另一如第7b圖之表示法，來掩飾業績下滑的事 實，使閱圖者不易察覺業績下滑的趨勢。前述台灣專利公 開第200632768號案之第4c圖也具有類似之圖形表示 法。另一方面，當業績表現平平時，如第8a圖所示，可 以藉由調整刻度範圍，如第8b圖所示，來操弄閱圖者對 圖形資料的感受。 【發明內容】 因此，本案之目的在於提供一種應用程式，其能將自 資料庫所取得之資料自動彙總，並運用實際値表示法、正 規化表示法、或等比例表示法，使資料使用者可以直接以 圖，對資料作出最正確的判斷與分析。 再者，本案之另一目的在於提供一套作圖方法，其能 將彙總後的資料，作更真實反映，使得閱圖者可以迅速看 出資料之趨勢、異常、或重大變動，不會爲作圖者操弄對 圖形資料的感受。 依據本發明之數値資料圖表化的方法，包含步驟··收 集數値資料；資料彙總，根據資料定義建立一支援資料處 -5- (3) (3)200905490 理的整合性平台’以對基本資料及各項資料數據作完整的 記錄；選取資料，針對彙總資料進行選擇；處理資料，以 ί守到衣圖所而r之要件；及圖表顯示資料，以所得到之製圖 要件進行製圖顯示。 【實施方式】 參考第1圖’其係爲本發明較佳實施例之將數値資料 圖表化的方法之流程圖。如第1圖所示’本發明之流程包 含.收集數値資料（s Τ 1 0);資料彙總（s Τ 2 0);選取資料 (ST3 0);處理資料（ST40);及圖表顯示資料（ST5〇)。 該步驟S T 1 0中之收集數値資料可以包含任何統計資 料表，例如’公司之資產負債表、損益表、公司內部各業 務員之銷售業績、顧客資料等等。 再者’進fT至資料囊總（S T 2 0)。先根據資料定義建立 一支援資料處理的整合性平台，以對基本資料及各項資料 數據作完整的記錄。在資料彙總的資料庫建置開發工具 上’本發明採用Visual Basic V6.0建立一套電子資料。 但本發明也可以使用其他程式語言，例如Excel，來建立 此電子資料，以將各種基本資料的欄位及項目與各發生時 間作各種的處理等，以利下一步驟（ST30)之輸入欄位資料 項目的選取。 再者，進行步驟ST30之選取資料，乃是針對彙總資 料進行選擇，可以例如選擇銷貨收入淨額、營業毛利等等 項目。 -6 - (4) (4)200905490 接著，進行步驟ST40之處理資料，以找出或計算出 製作想要圖形的縱軸上、下限及其他製圖要件。 最後，以找出或計算出的縱軸上、下限及其他製圖要 件製作圖形顯示（S T 5 0 )。 另外，將針對處理資料步驟ST40的步驟作進一步的 說明。如第2圖所示，本發明之資料處理方法被分成三種 類型，分別歸類爲實際値表示法、正規化表示法及等比例 表示法’並各自以第一、第二與第三實施例加以說明。 [第一實施例] 實際値表示法 本發明第一實施例係爲以實際値作爲圖形的縱座標刻 度，其處理資料方法請參考第3圖說明。該流程開始於 ST410，然後，在步驟ST41 1所輸入資料項中找出最大値 及最小値做爲上、下限。接著，在步驟S T4 1 2中，將最 大値與最小値相減，所得再行除以想要標示的資料數目， 即可求得軸距，例如當資料顯示爲紙張或電腦監視器時， 則可以除以1 〇，反之，如果在手持式裝置，例如PDA及 手機等之顯示裝置時，則可以除以5，以求得適當的軸 距。另外，在步驟ST4 1 3中，如果使用者想保有圖形之邊 寬，則可以依想要的邊寬，將最大値及最小値加上1倍或 半倍的軸距，作爲資料顯示或列印。透過以上步驟，便可 以在步驟ST414得到圖形縱軸的新上限及下限，當然，若 是閱圖者並未選擇加上邊寬，則新上、下限將等於原本所 (5) (5)200905490 選擇之上、下限。接下來，在步驟ST415中，也可以將各 輸入資料總和除以有效的資料數目，而得到平均値線。當 然，也可以將 〇參考線直接繪出，作爲閱圖參考 (ST416)。最後’此流程結束（ST417)。 雖然’此資料處理步驟ST410係以包含有找出最大及 最小値做爲上、下限（S T4 1 1)、求出軸距（S T4 1 2)、加邊寬 (ST413)、求得新上、下限（ST414)、求平均値線 (ST415)、求0參考線（ST416)及結束（ST417)等步驟加以 說明，但明顯地，本發明之方法的步驟ST412、ST413、 ST4 14、ST4 15、ST416並非全然必要，也可以只執行步驟 ST4 1 0、ST4 1 1、ST4 1 7 即可。 因此，假設在單縱軸（單數列）下，在1至1 0點中， 分別有數値： 85 76 79 86 8 1 77 7 1 76 82 73 則以EXCEL自動作圖表示出如第9a圖所示之線形 圖，但此線形圖之資料振幅並非最大，可能使得閱圖者， 無法察覺其中之變化情形。此資料在經過本發明取資料中 之最大値與最小値之資料處理後，得到如第9b圖所示之 線形圖。如第9b圖所示，本發明所繪製之線形圖的上、 下限爲數列之最大、最小値，同時，其間之軸距也不再是 整數。相較於第9a圖之線形圖，本發明之實際値表示法 更能確切反映資料起伏變動的情形，使得閱圖者容易看出 -8- (6) (6)200905490 資料的特性與異常。 此外，假設在單縱軸（單數列）下，在1至1 〇點中， 分別有數値： 45 -28 3 0 40 50 30 39 -36 56 則以EXCEL自動作圖表示出如第10a圖所示之線形 圖。而此資料在經過本發明取資料中之最大値與最小値之 處理資料方法後，得到如第1 Ob圖所示之線形圖。相較於 第1 圖之線形圖，本發明實際値表示法之線形圖，更輔 以平均値線與0參考線，使閱圖者更容易了解到資料之集 中趨勢及是否變號。 再者’當資料爲如下之數字資料之雙數列： 數列1 65 84 61 81 60 75 69 64 71 69 數列2 105 115 101 111 96 115 113 104 111 109 當資料以EXCEL自動作出線形圖時，會有如第n a 圖所不之結果。而當資料經本發明之實際値表示法後，此 兩資料組中之上限與下限爲分別丨i 5及6〇，其軸距爲 5.5 °因此’完成了如第1 lb圖所示之線形圖。此第nb 圖之資料線形明顯較第1 1 a圖更能反映資料的劇烈變動。 當然’本發明也可以以條形圖來表示，相同於第n b 圖之資料也可以表示成第1 1 c圖所示。 -9- (7) (7)200905490 [第二實施例] 正規化表示法 本實施例之正規化表示法的處理資料方法請參考第4 圖說明。該流程開始於ST420，然後在步驟ST421將該筆 資料中之所有資料讀入，然後，在步驟ST422中，將各資 料取絕對値。接著，在步驟ST423中，將絕對値中之最大 者作爲正規化因數，將各資料--除以該正規化因數，如 此即取得正規化，因此’所有資料均會被正規化。接著， 在步驟ST424中，進行取上、下限步驟。由於經正規化後 必定有一 +1或-1，所以，若是閱圖者想要清楚地看到資 料之趨勢與變動，則可以指定+ 1作爲最大値（上限），而最 小値作爲下限，或者，指定-1作爲最小値（下限）而最大値 作爲上限，當然’若是閱圖者想要看出資料之散佈範圍及 變號情況，則可以直接將上限與下限指定爲1與-1。因 此，閱圖者將可以看出資料之散佈範圍及變號情況。接 者’執行軸距的計算（ST42S)。另外，在步驟ST426中， 如果使用者想保有線形圖之邊寬，則可以依想要的邊寬， 將最大値及最小値加上1倍或半倍的軸距，作爲資料顯示 或列印。透過以上步驟，便可以在'步驟ST427得到圖形縱 軸的新上限及下限，當然，若是閱圖者並未選擇加上邊 寬，則新上、下限將等於原本所選擇之上、下限。最後， 此流程結束（S T 4 2 8 )。 雖然’此資料處理步驟ST420係包含有ST421、 -10- (8) (8)200905490 ST422 ' ST423 、 ST424 、 ST425 、 ST426 、 ST427 及 ST428 等步驟加以說明，但明顯地，本發明之方法的步驟 ST425、ST426、ST427並非全然必要，也可以只執行步驟 ST420 、 ST421 、 ST422 、 ST423 、 ST424 、 ST428 即可。 以下將說明實際値表示法與正規化表示法之各自的用 途，假設在單縱軸下，在1至1 〇點中，分別有數列1、 數列2、數列3、數列4、數列5 : 數列1 1 2 3 4 4.1 3 1.5 1 5 1 數列2 23 24 25 26 26.1 25 23.5 23 24.5 23 數列3 991 438 522 -232 -377 214 111 273 -399 459 數列4 -12 -24 -22 -46 -19 -22 -35 -19 -62 -42 數列5 -649 -499 -702 -442 -231 -290 -718 -561 -233 -700 由於數列資料之間的差異很大，導致不論是以 EXCEL自動作圖（如圖第12a圖）或以實際値表示法之資料 處理方法顯示時（如第1 2b圖）’皆無法讓閱圖者看出數列 1、數列2、數列4之趨勢及變動。但是，再運用正規化 表示法之資料處理方法後卻可以得到第1 2c圖’一個可以 清楚看到五組數列之趨勢與變化的圖。 此第1 2c圖的意義，可以是當例如數列2爲營業額’ 而數列1可以是獲利率，由於，兩數列之實質數値差異很 大，以實際値表示法，也不太能看出此兩數列之關係’透 過本正規化表示法’可以使閱圖者輕易了解到營業額與獲 -11 - (9) (9)200905490 利率之關係，以利閱圖者作出立即之資料判斷。 當然，若是閱圖者想要清楚地看到資料之趨勢與變 動’則可以指定+ 1作爲最大値（上限），而最小値作爲下 限’或者’指定-1作爲最小値（下限）而最大値作爲上限。 例如，在單縱軸下’在1至1 0點中，分別有數列1、數 列2、數列3、數列4、數列5，其數値如下： 數列 1 3.5 2.4 3 4 5 3 4.1 4.4 4.2 2.8 數列 2 25.5 24.4 27 26 22 25 26.1 22.3 23 24.8 數列3 99 1 43 8 522 732 5 77 5 14 76 1 473 699 459 數列 4 52 4 1 42 46 39 3 9 35 33 62 42 數列 5 649 499 702 442 3 3 1 3 99 7 18 56 1 433 700 運用EXCEL自動作圖（第13a圖），依然無法顯示數 列1、數列2、數列4之變動情況。但若是運用本發明之 正規化表示法找出數列之最大、最小値做爲上、下限，則 可以得到第1 3 b圖之顯示結果，其資料係在+ 1與最小値 (0.441978)間變動’看出資料的趨勢及變動情形；反 之’若閱圖者想要看到的是資料散佈範圍及變號情況，則 可以強制「設定上限爲1、下限爲-1」之方法，如此一來 可以得到1 3 c圖，看出此五組數列皆爲正，且其皆散佈於 圖形的某一端，表示這幾組數列之變動皆不大且沒有變 號。 -12- (10) 200905490 [第三實施例] 等比例表示法 ' 當兩資料組要在雙縱軸上做比較時，由於傳統的作圖 .的方法並沒有一個準則，常常使得比較的兩縱軸之顯示比 例尺不一致而誤導了閱圖人的感受，在此情況下，本發明 等比例表示法之資料處理流程更能顯現其功效。本實施例 之等比例表示法的處理資料流程請參考第5圖說明。該流 程開始於ST430，然後在ST431找出兩資料組中不需要進 行修正之最大、最小値做爲上、下限，接著在ST43 2進行 運算，求得需修正縱軸之上限或下限、求兩邊縱軸之軸距 (ST43 3 )、加上邊寬（ST434)、求得縱軸之新上、下限 (ST43 5 )，最後，結束流程（ST43 6)。 雖然，此資料處理步驟ST43 0係以包含有找出不需要 進行修正之最大、最小値做爲上、下限（ST431)、進行運 算求得需修正軸修正後之上限或下限（ST43 2)、求得兩邊 縱軸之軸距（ST43 3)、加上邊寬（ST434)、求得縱軸之新 上、下限（ST43 5)、結束流程（ST436)等步驟加以說明， 但明顯地，本發明之方法的步驟 ST43 3、ST434、ST435 並非全然必要，也可以只執行步驟 ST43 0、ST431、 ST432 ' ST436 即可。 其中，步驟ST431與ST43 2將在第6圖做進一步的 說明。 首先，在步驟ST601中，分別求得要比較之兩數列之 E値，其方法爲將個別數列中之最大値減去數列中之最小 -13- (11) (11)200905490 値後再除以數列中所有資料取絕對値後之最大値。然後， 在步驟ST602中，將兩資料組的E値相比，則具有較大E 値之資料組爲第一資料組，而較小E値者爲第二資料組， 而當E値相等時，則令任一資料組爲第一資料組。 如第6圖中所示，在第一資料組之E値大於等於第二 資料組之E値之情況下，第一資料組維持以其最大、最小 値做爲其縱軸之上、下限，而第二資料組則需要變動其縱 軸之上限或下限以使得兩數列之兩縱軸之顯示比例尺相 同。 其中’在第一類情況下，也就是在第一資料組之最大 値取絕對値後大於等於第一資料組之最小値取絕對値、第 二資料組之最大値取絕對値後大於等於第二資料組之最小 値取絕對値的情況下，則在步驟S T 6 1中，第二資料組之 下限需要調整（也就是第二資料組之上限不需要調整），以 使得兩縱軸之顯示比例尺相同，其方法爲在步驟ST611 中，第二資料組上限χ(第一資料組之下限+第一資料組上 限）。 在第二類情況下，也就是在第一資料組之最大値取絕 對値後大於等於第一資料組之最小値取絕對値、第二資料 組之最大値取絕對値後小於第二資料組之最小値取絕對値 的情況下，則在步驟ST62中，第二資料組之上限需要調 整（也就是第二資料組之下限不需要調整），以使得兩縱軸 之顯示比例尺相同，其方法爲在步驟ST621中，第二資料 組下限X (第一資料組之下限+第一資料組上限）。 -14- (12) (12)200905490 在第三類情況下’也就是在第一資料組之最大値取絕 對値後小於第一資料組之最小値取絕對値、第二資料組之 最大値取絕對値後大於等於第二資料組之最小値取絕對 値’則在步驟ST63中’第二資料組之下限需要調整（也就 是第二資料組之上限不需要調整），以使得兩縱軸之顯示 比例尺相同’其方法爲在步驟S T 6 3 1中，第二資料組上限 x(第一資料組之上限+第一資料組下限）。 在第四類情況下’也就是在第一資料組之最大値取絕 對値後小於第一資料組之最小値取絕對値 '第二資料組之 最大値取絕對値後小於第二資料組之最小値取絕對値的情 況下，則在步驟ST64中’第二資料組之上限需要調整（也 就是第二資料組之下限不需要調整），以使得兩縱軸之顯 示比例尺相同，其方法爲在步驟S T 6 4 1中，第二資料組下 限X (第一資料組之上限+第二資料組下限）。 例如，現在有A、B兩組數列要做雙縱軸比較： A數列 5941 •6382 7209 -5323 7775 -3142 6117 5737 -4657 7122 B數列 72 84 92 96 71 82 85 69 76 82 當A、B數列以EXCEL自動作出雙縱軸圖時，會得 到如第14a圖之顯示結果。然而，由14b與14c圖（皆由 EXCEL自動作圖）可以發現，EXCEL所做出之雙縱軸圖只 是將原本的B數列單縱軸自動作圖之左縱軸移到雙縱軸之 右縱軸，由於兩軸之顯示比例尺並不相同，將此兩數列放 在同一線形圖上’無法公平互相比較，並且將存在：1·圖 -15- (13) (13)200905490 形依然顯示比例尺非最大，閱圖者不容易察覺其中趨勢、 異常與重大變動之情況2.未考慮雙縱軸之間的比例關係， 使得閱圖人容易誤解兩數列之變動幅度，而此比例問題， 更可能給予作圖人操控閱圖人感受的機會。 因此，需要透過本發明等比例表示法之資料處理流程 給予雙數列同比例的縱軸。由於 A數列之 E値爲 1.984823、B數列之E値爲0.28125，因此，A數列爲第 一資料組，A數列之最大、最小値依然爲其縱軸之上、下 限，而B數列則需要變動其縱軸之上限或下限以使得兩數 列之兩縱軸之顯示比例尺相同。 由於A數列之最大値取絕對値後大於A數列之最小 値取絕對値，且B數列之最大値取絕對値後大於B數列 之最小値取絕對値（屬於第一類情況），因此，B數列縱軸 之下限需要向下修正以達到雙方的顯示比例尺相同，其方 法爲B數列上限x(A數列之下限+A數列之上限）。 經過本發明等比例表示法之資料處理方法調整後可以 得到，A數列之上限爲7775、A數列之下限爲_63 82、b 數列之上限爲96、B數列之下限爲- 78.80025，如第I4d 圖所示。透過此方法可以得到兩點最大的功效：1 ·由於此 兩筆資料縱軸顯示比例尺相等’資料間可以作更公平的比 例顯示。2.在兩數列顯示比例尺相等的情況下，將正規化 後之數値回歸到其原本之單位數値，使閱圖者更容易了解 原資料之數値大小。 上述之公平性可以透過將A、B數列正規化後之數値 -16- (14) (14)200905490 及運用正規化表示法所作出之第14e、14f圖與第14d圖 比較來驗證。A數列及B數列正規化後之數値如下： A數列 0.7641 -0.8208 0.9272 -0.6846 1 -0.4041 0.7867 0.7378 -0.5989 0.91601 B數列 0.75 0.875 0.9583 1 0.7395 0.8541 0.8854 0.7187 0.7916 0.8541 由上表可以看到正規化後，A數列之最大値爲1、最 小値爲-〇 . 8 2 0 8 4、B數列之最大値爲 1、最小値爲 0.7 1 87 5。將A、B數列正規化後之最大、最小値各做爲雙 縱軸圖左軸及右軸之上、下限可以發現兩者之縱軸顯示比 例尺完全不同，如此比較並不合理，如第1 4e圖所示；因 此必須要將縱軸顯示比例尺較小的一方（也就是以B數列 之最大、最小値爲上、下限的右縱軸），重新調整爲與縱 軸顯示比例尺較大的一方相等（也就是以A數列之最大、 最小値爲上、下限的左縱軸），如此一來便可以達到在兩 者之軸距比例相同的條件下以最大的拉撐幅度幫助閱圖者 最容易察覺數値之趨勢與變動，如第14f圖所示。我們可 以發現，第14f圖與等比例表示法所做出來之圖（第14d 圖）拉撐幅度一模一樣，可見得等比例表示法之公平性， 唯等比例表示法所做出之圖可以使兩數列回歸到原本之數 値，幫助閱圖者更容易察覺數列之實際大小。 當然，本發明也可以以線形圖搭配條形圖來表示，相 同於第14d圖之資料也可以表示成第14g圖所示（上、下 各加一倍軸距邊寬）。 -17- (15) (15)200905490 又例如，現在有C、D兩組數列要做雙縱軸比較·· C數列 72 84 92 96 71 82 85 69 76 82 D數列 -1230 -4245 -2249 -2654 -3234 -3790 -4593 -1983 -980 -789 當C、D數列以EXCEL自動作出雙縱軸圖時，會得 到如第15a圖之顯示結果。然而，由15b與15c圖（皆由 EXCEL自動作圖）可以發現，同樣的，EXCEL所做出之雙 縱軸圖是將原本的D數列單縱軸自動作圖之左縱軸移到 雙縱軸之右縱軸，由於兩軸之顯示比例尺並不相同，將此 兩數列放在同一線形圖上，無法公平地互相比較，並且將 存在：1 .圖形依然顯示比例尺非最大，閱圖者不容易察覺 其中趨勢、異常與重大變動之情況2.未考慮雙縱軸之間的 比例關係，使得閱圖人容易誤解兩數列之變動幅度’而此 比例問題，更可能給予作圖人操控閱圖人感受的機會。 因此，需要透過本發明等比例表示法之資料處理流程 給予雙數列同比例的縱軸。由於C數列之E値爲 0.28125、D數列之E値爲0.828217’因此’ D數列爲第 一資料組，D數列之最大、最小値依然爲其縱軸之上、下 限，而C數列則需要變動其縱軸之上限或下限以使得兩數 列之兩縱軸之顯示比例尺相同。 由於D數列之最大値取絕對値後小於D數列之最小 値取絕對値，且C數列之最大値取絕對値後大於c數列 之最小値取絕對値（屬於第三類情況），因此’ C數列縱軸 -18- (16) (16)200905490 之下限需要向下修正以達到雙方的顯示比例尺相同，其方 法爲C數列上限X(D數列之上限—D數列之下限）。 經過本發明之等比例表示法調整後可以得到，C數列 之上限爲96、C數列之下限爲16.491 18、D數列之上限 爲-7 8 9、D數列之下限爲-4 5 9 3，如第1 5 d圖所示。同樣 的，透過此方法可以得到兩點最大的功效：1 .由於此兩筆 資料顯示比例尺相同，資料間可以作更公平的比例顯示。 2 ·在兩數列之縱軸顯示比例尺相等的條件下，將正規化後 之數値回歸到其原本之單位數値，使閱圖者更容易了解原 資料之數値大小。 上述之公平性也可以透過將C、D數列正規化後之數 値及運用正規化表示法所作出之第15e、15f圖與第15d 圖比較來驗證。C數列及D數列正規化後之數値如下： C數列 0.75 0.875 0.9583 1 0.7395 0.8541 0.8854 0.7187 0.7916 0.8541 D數列 -0.2678 -0.9242 -0.4896 -0.5778 -0.7041 -0.8251 -1 -0.4317 -0.2133 -0.1717 由上表可以看到正規化後，C數列之最大値爲1、最 小値爲0.7 1 8 7 5、D數列之最大値爲_0.1 7 1 784、最小値爲- 1。將C、D數列正規化後之最大、最小値各做爲雙縱軸 圖左軸及右軸之上、下限可以發現兩者之軸距顯示比例尺 完全不同，如此比較並不合理，如第1 5 e圖所示；因此必 須要將軸距顯示比例尺較小的一方（也就是以C數列之最 大、最小値爲上、下限的左縱軸），重新調整爲與縱軸顯 -19- (17) (17)200905490 示比例尺較大的一方相等（也就是以D數列之最大、最小 値爲上下限的右縱軸）’如此一來便可以達到在兩者之縱 軸比例相同的條件下以最大的拉撐幅度幫助閱圖者最容易 察覺數値之趨勢與變動’如第15f圖所示。我們可以發 現，第15f圖與等比例表示法所做出來之圖（第15 d圖）拉 撐幅度一模一樣’可見得等比例表示法之公平性’唯等比 例表示法所做出之圖可以使兩數列回歸到原本之數値’幫 助閱圖者更容易察覺原資料之數値大小。 當然，本發明也可以以線形圖搭配條形圖來表示’相 同於第15d圖之資料也可以表示成第15g圖所示（上、下 各加一倍軸距邊寬）。 綜上所陳，本案係一種有關於將數値資料圖表化的方 法，其能將大量資料加以彙總分析’並根據閱圖者之不同 需求及資料之筆數，運用本案所揭示實際値表示法、正規 化表示法、或等比例表示法之不同資料處理方法以利閱圖 者能更明確得知資料的實質意義，而不會被作圖者所操 控。 【圖式簡單說明】 第1圖爲一流程圖，用以解釋將數値資料圖表化的步 驟； 第2圖爲一示意圖，用以例示第1圖中處理資料步驟 之進一步細節； 第3圖爲一流程圖，用以解釋實際値表示法之處理資 -20- (18) (18)200905490 料步驟； 第4圖爲一流程圖，用以解釋正規化表兩法之處理資 料步驟； 第5圖爲一流程圖，用以解釋等比例表示法之處理資 料步驟； 第6圖爲一流程圖，用以解釋等比例表不法之求得修 正後上、下限步驟之進一步細節； 第7a、7b、8a、8b圖用以說明不同之作圖方法會帶 給閱圖者完全不同感受之實例； 第9a及9b圖分別爲運用傳統方式與本發明之實際値 表示法所得之單縱軸單數列線形圖； 第10a及10b圖分別爲運用傳統方式與本發明之實際 値表示法加上平均値線與〇參考線所得之單縱軸單數列線 形圖： 第11a、lib圖分別爲運用傳統方式、本發明之實際 値表示法所得之單縱軸雙數列線形圖； 第11c圖爲運用本發明之實際値表示法所得之單縱軸 雙數列條形圖； 第12a、12b、12c圖分別爲運用傳統之作圖方法、本 發明之實際値表示法、本發明之正規化表示法所得之單縱 軸多數列線形圖； 第13a、13b、13c圖分別爲運用傳統之作圖方法、本 發明之找出數列之最大、最小値做爲上、下限之正規化表 示法、本發明之設定上限爲1、下限爲之正規化表示法 -21 - (19) (19)200905490 所得之單縱軸多數列線形圖； 第i 4a、14b、14c圖分別爲運用傳統作圖方法所得之 A、B數列之雙縱軸圖、A數列之單縱軸圖、B數列之單 縱軸線形圖； 第1 4d圖則爲運用本發明之等比例表示法所得之A、 B數列之雙縱軸線形圖； 第1 4 e、1 4 f圖分別爲運用正規化後之最大、最小値 分別作爲兩縱軸之上、下限之圖及將正規化後之雙縱軸等 比例修正後所得之線形圖； 第1 4 g圖爲運用本發明之等比例表示法所得之A、B 數列之雙縱軸線形搭配條形圖； 第1 5 a、1 5 b、1 5 c圖分別爲運用傳統作圖方法所得之 C、D數列之雙縱軸圖、C數列之單縱軸圖、D數列之單 縱軸圖； 第1 5 d圖則爲運用本發明之等比例表示法所得之c、 D數列之雙縱軸線形圖； 第1 5 e、1 5 f圖分別爲運用正規化後之最大、最小値 分別作爲兩縱軸之上 '下限之圖及將正規化後之雙縱軸等 比例修正後所得之線形圖；及 第15g圖爲運用本發明之等比例表示法所得之C、D 數列之雙縱軸線形搭配條形圖。 -22-200905490 (1) IX. Description of the invention [Technical field to which the invention pertains] The present invention relates to a method of graphing data, and more specifically, the present invention relates to the use of information, particularly in financial statements. 'A method of trending, anomalous, or significant change in the general ledger. [Prior Art] In the current related database, the data is managed in rows and columns. 'There is only the relationship between the field and the field, or the data grid and the data grid, but the relationship and sequence are lacking. Sex, the data integration required for analysis, must rely on the human brain to make judgments. If you want to compare with last year or last month, you need to write S Q L (Structured Query Language) to do it. It is not only troublesome, but also often delays the timeliness of decision-making. Therefore, there are OLAP (Online Analytical Processing) analysis tools on the market that can efficiently classify data according to dimensions in advance, and more importantly, it can establish a so-called time dimension, which is to classify data by time. The cutting of the time dimension can be different according to the data attributes, such as the shipment volume in days, and the higher-level reports, such as the sales trend, and the earnings, in quarters. Therefore, in the case of Taiwan Patent Publication No. 20063276 No. 8 , it discloses 'using a OLAP to conduct a legal investigation of a database containing financial information' so that the database user can easily obtain the amount of data for a specific period, and then proceed Mutual comparison or mastering the change of trend 'to achieve -4- (2) (2) 200905490 fast best analysis results. The case still follows the majority of today's managers who use EXCEL or similar mapping tools for graphical performance. However, the graphical representation of excel is still the case of traditional statistical charts and viewers. For example, As for the sales performance of the 7th to the 7th of the 7th chart, the curve of the chart first climbs up to the highest point and then goes down. Therefore, the mapper will want to express it as another figure 7b. The law, to cover up the fact of the decline in performance, makes it difficult for readers to perceive the downward trend in performance. The 4th figure of the aforementioned Taiwan Patent Publication No. 200632768 also has a similar graphical representation. On the other hand, when the performance is flat, as shown in Figure 8a, the viewer's perception of the graphic material can be manipulated by adjusting the scale range, as shown in Figure 8b. SUMMARY OF THE INVENTION Accordingly, the object of the present invention is to provide an application that automatically aggregates data obtained from a database and uses actual 値 representation, normalized representation, or proportional representation to enable data users. The most accurate judgment and analysis of the data can be made directly by the map. Furthermore, another purpose of the case is to provide a set of mapping methods that can reflect the aggregated data more accurately, so that the viewer can quickly see the trend, abnormality, or major changes of the data. The mapper manipulates the feelings of the graphic material. The method for charting data according to the present invention includes the steps of collecting data; collecting data, and establishing a supporting information office according to the definition of the data - 5 - (3) (3) 200905490 The integrated platform of the right The basic data and various data are completely recorded; the data is selected, the summary data is selected; the data is processed, and the data is displayed in the form of the drawing; and the chart is displayed, and the drawing is displayed in the drawing. . [Embodiment] Referring to Fig. 1 which is a flow chart of a method for charting data according to a preferred embodiment of the present invention. As shown in Fig. 1, the process of the present invention includes: collecting data (s Τ 1 0); data summary (s Τ 2 0); selecting data (ST3 0); processing data (ST40); (ST5〇). The collected data in this step S T 1 0 can contain any statistical data table, such as 'company's balance sheet, profit and loss statement, sales performance of various business operators within the company, customer information, and the like. Furthermore, 'into fT to the total data capsule (S T 2 0). First, an integrated platform for supporting data processing will be established based on the data definition to make a complete record of the basic data and various data. In the data collection database development tool, the invention uses Visual Basic V6.0 to create a set of electronic materials. However, the present invention can also use other programming languages, such as Excel, to establish the electronic material, to perform various processing of fields and items of various basic materials and each occurrence time, so as to facilitate the input field of the next step (ST30). Selection of data items. Further, the selection of the information in step ST30 is performed for the summary information, and for example, the net sales revenue, the operating margin, and the like can be selected. -6 - (4) (4) 200905490 Next, the processing data of step ST40 is performed to find out or calculate the upper and lower limits of the vertical axis and other drawing requirements for the desired pattern. Finally, a graphical display (S T 5 0 ) is created by finding or calculating the upper and lower limits of the vertical axis and other drawing requirements. Further, the steps of the processing data step ST40 will be further explained. As shown in FIG. 2, the data processing method of the present invention is divided into three types, which are respectively classified into an actual 値 representation, a normalized representation, and an equal-scale representation, and are each implemented in the first, second, and third embodiments. An example is given. [First Embodiment] Actual 値 Representation The first embodiment of the present invention is an ordinate scale with an actual 値 as a figure, and the method of processing data is described with reference to Fig. 3. The flow starts at ST410, and then, in the data item input in step ST41 1, the maximum 値 and the minimum 找出 are found as the upper and lower limits. Then, in step S T4 12 2, the maximum 値 is subtracted from the minimum ,, and the obtained re-row is divided by the number of data to be marked, and the wheelbase can be obtained, for example, when the data is displayed as paper or a computer monitor. Then, it can be divided by 1 〇. Conversely, if it is a display device of a handheld device such as a PDA or a mobile phone, it can be divided by 5 to obtain an appropriate wheelbase. In addition, in step ST4 13 , if the user wants to keep the side width of the graphic, the maximum and minimum 値 can be added to the wheelbase of 1 or half times as the data display or column according to the desired side width. Printed. Through the above steps, the new upper and lower limits of the vertical axis of the graph can be obtained in step ST414. Of course, if the viewer does not choose to add the edge width, the new upper and lower limits will be equal to the original (5) (5) 200905490. Upper and lower limits. Next, in step ST415, the sum of the respective input data may be divided by the number of valid data to obtain an average line. Of course, you can also draw the 〇 reference line directly as a reference for reading (ST416). Finally 'this process ends (ST417). Although 'this data processing step ST410 includes to find the maximum and minimum 値 as the upper and lower limits (S T4 1 1), find the wheelbase (S T4 1 2), add the edge width (ST413), find new The steps of the upper and lower limits (ST414), the averaged enthalpy line (ST415), the zero reference line (ST416), and the end (ST417) are described, but obviously, steps ST412, ST413, ST4 14, and ST4 of the method of the present invention are apparent. 15. ST416 is not absolutely necessary, and only steps ST4 1 0, ST4 1 1 and ST4 1 7 may be performed. Therefore, it is assumed that under the single vertical axis (singular column), there are several numbers in the 1 to 1 0 point: 85 76 79 86 8 1 77 7 1 76 82 73 is automatically represented by EXCEL as shown in Figure 9a. The line graph is shown, but the amplitude of the data of this line graph is not the largest, which may make the viewer unable to detect the change. This data is processed by the data of the largest and smallest data in the data taken by the present invention to obtain a line graph as shown in Fig. 9b. As shown in Fig. 9b, the upper and lower limits of the line graph drawn by the present invention are the maximum and minimum 数 of the series, and the wheelbase therebetween is no longer an integer. Compared with the line graph of Fig. 9a, the actual 値 representation of the present invention can more accurately reflect the fluctuation of the data, so that the reader can easily see the characteristics and anomalies of the data of -8-(6)(6)200905490. In addition, it is assumed that under the single vertical axis (single-column), there are several numbers in the 1 to 1 〇 point: 45 -28 3 0 40 50 30 39 -36 56 is automatically represented by EXCEL as shown in Figure 10a. A line graph showing. And this data is obtained by the method of processing the data of the maximum 値 and the minimum 取 in the data of the present invention, and the line graph as shown in the first Ob is obtained. Compared with the line graph of Fig. 1, the line graph of the actual 値 representation of the present invention is supplemented by the average 値 line and the 0 reference line, which makes it easier for the reader to understand the trend of the data and whether it is changed. Furthermore, 'when the data is a double column of the following digital data: series 1 65 84 61 81 60 75 69 64 71 69 series 2 105 115 101 111 96 115 113 104 111 109 When the data is automatically made by EXCEL, there will be The result of the na diagram is not. When the data is represented by the actual 値 representation of the present invention, the upper and lower limits of the two data sets are 丨i 5 and 6〇, respectively, and the wheelbase is 5.5 °, thus completing the line graph as shown in Figure 1 lb. . The data line shape of this nb graph is significantly more reflective than the 1st 1 a map. Of course, the present invention can also be represented by a bar graph, and the data similar to the nth graph can also be represented as the 1st 1c diagram. -9- (7) (7) 200905490 [Second embodiment] Normalized representation The method of processing data of the normalized representation of this embodiment is described with reference to FIG. The flow starts at ST420, and then all the data in the pen data is read in step ST421, and then, in step ST422, the respective materials are taken absolutely. Next, in step ST423, the largest one of the absolute 値 is used as the normalization factor, and the data is divided by the normalization factor, and thus the normalization is performed, so that all the data are normalized. Next, in step ST424, the steps of taking the upper and lower limits are performed. Since there must be a +1 or -1 after normalization, if the reader wants to clearly see the trend and change of the data, you can specify + 1 as the maximum 上限 (upper limit), and the minimum 値 as the lower limit, or Specify -1 as the minimum 下限 (lower limit) and maximum 値 as the upper limit. Of course, if the reader wants to see the distribution range and the variable number of the data, the upper and lower limits can be directly specified as 1 and -1. Therefore, the reader will be able to see the distribution of the data and the change of the number. The receiver performs the calculation of the wheelbase (ST42S). In addition, in step ST426, if the user wants to keep the side width of the line graph, the maximum and minimum 値 plus 1 or half times the wheelbase can be displayed or printed as data according to the desired side width. . Through the above steps, the new upper and lower limits of the vertical axis of the graph can be obtained in step ST427. Of course, if the viewer does not select the edge width, the new upper and lower limits will be equal to the upper and lower limits selected. Finally, the process ends (S T 4 2 8 ). Although 'this data processing step ST420 includes the steps of ST421, -10-(8) (8) 200905490 ST422 'ST423, ST424, ST425, ST426, ST427 and ST428, it is obvious that the steps of the method of the present invention are obvious. ST425, ST426, and ST427 are not all necessary. You can also perform only steps ST420, ST421, ST422, ST423, ST424, and ST428. The respective uses of the actual 値 representation and the normalized representation will be described below. It is assumed that under the single vertical axis, there are several columns, one, two columns, two columns, three columns, and five columns in the number of 1 to 1 :: series 1 1 2 3 4 4.1 3 1.5 1 5 1 Sequence 2 23 24 25 26 26.1 25 23.5 23 24.5 23 Sequence 3 991 438 522 -232 -377 214 111 273 -399 459 Sequence 4 -12 -24 -22 -46 -19 -22 -35 -19 -62 -42 Sequence 5 -649 -499 -702 -442 -231 -290 -718 -561 -233 -700 Due to the large difference between the series of data, it is automatically drawn by EXCEL. (as shown in Figure 12a) or when the data processing method of the actual 値 representation is displayed (such as Figure 12b), it is impossible for the reader to see the trend and variation of the series 1, column 2, and column 4. However, by using the data processing method of the normalized representation method, it is possible to obtain a map of the 1st and 2nd graphs, which can clearly see the trend and change of the five groups of numbers. The meaning of this 12c picture can be that, for example, the number 2 is the turnover' and the number 1 can be the interest rate. Since the difference between the two numbers is very large, it can not be seen in the actual 値 representation. The relationship between the two series 'through this formalized representation' can make the readers easily understand the relationship between the turnover and the interest rate of -11 - (9) (9) 200905490, in order to facilitate the reader to make an immediate judgment. . Of course, if the reader wants to clearly see the trend and change of the data, then you can specify + 1 as the maximum 上限 (upper limit), and the minimum 値 as the lower limit ' or 'specify -1 as the minimum 下限 (lower limit) and the maximum 値As the upper limit. For example, in the single vertical axis, 'from 1 to 10 o'clock, there are several columns, one column, two columns, three columns, four columns, and five columns, the number of which is as follows: Series 1 3.5 2.4 3 4 5 3 4.1 4.4 4.2 2.8 Series 2 25.5 24.4 27 26 22 25 26.1 22.3 23 24.8 Sequence 3 99 1 43 8 522 732 5 77 5 14 76 1 473 699 459 Sequence 4 52 4 1 42 46 39 3 9 35 33 62 42 Sequence 5 649 499 702 442 3 3 1 3 99 7 18 56 1 433 700 With EXCEL automatic drawing (Fig. 13a), the changes in the series 1, column 2, and column 4 are still not displayed. However, if the maximum and minimum 数 of the series are used to find the upper and lower limits by using the normalized representation of the present invention, the display result of the 1 3 b graph can be obtained, and the data is changed between + 1 and the minimum 値 (0.441978). 'See the trend and changes of the data; on the other hand, if the reader wants to see the data distribution range and the variable number, you can force the method of setting the upper limit to 1, the lower limit to -1. You can get the 1 3 c graph. It can be seen that the five groups of numbers are all positive, and they are all scattered at one end of the graph, indicating that the changes in these groups of numbers are not large and there is no change. -12- (10) 200905490 [Third Embodiment] Isometric Representation' When two data sets are to be compared on the double vertical axis, since the traditional method of drawing does not have a criterion, the two are often compared. The display scale of the vertical axis is inconsistent and misleads the viewer's feelings. In this case, the data processing flow of the proportional representation method of the present invention can more clearly show its efficacy. For the processing data flow of the proportional representation method of this embodiment, please refer to FIG. The process starts at ST430, and then in ST431, the maximum and minimum 不需要 of the two data sets that do not need to be corrected are determined as the upper and lower limits, and then the operation is performed in ST43 2, and the upper or lower limit of the vertical axis needs to be corrected, and both sides are sought. The wheelbase of the vertical axis (ST43 3 ), the side width (ST434), and the new upper and lower limits of the vertical axis are obtained (ST43 5 ), and finally, the flow is terminated (ST43 6). In the data processing step ST43 0, the upper limit or the lower limit (ST43 2) after the correction of the axis is corrected by including the maximum and minimum values that do not need to be corrected as the upper and lower limits (ST431), and the calculation is performed. The steps of obtaining the wheelbase (ST43 3) of the vertical axes on both sides, adding the side width (ST434), finding the new upper and lower limits of the vertical axis (ST43 5), and ending the flow (ST436) are explained, but obviously, the present invention The steps ST43 3, ST434, and ST435 of the method are not absolutely necessary, and only steps ST43 0, ST431, and ST432 'ST436 may be executed. Among them, steps ST431 and ST43 2 will be further explained in Fig. 6. First, in step ST601, the E値 of the two series to be compared are respectively obtained by subtracting the largest 个别 in the individual series from the smallest -13(11)(11)200905490 in the series, and then dividing by The maximum number of all data in the series is taken as absolute. Then, in step ST602, comparing the E値 of the two data sets, the data group having a larger E 为 is the first data group, and the smaller E 値 is the second data group, and when E 値 is equal, , then make any data group the first data group. As shown in Fig. 6, in the case where the E値 of the first data set is greater than or equal to the E値 of the second data set, the first data set maintains its maximum and minimum 値 as its upper and lower limits, The second data set needs to change the upper or lower limit of its vertical axis so that the display scales of the two longitudinal axes of the two series are the same. Wherein 'in the first case, that is, after the absolute maximum 第一 of the first data set is greater than or equal to the minimum of the first data set, the absolute maximum 値, the maximum of the second data set is greater than or equal to the first If the minimum of the second data set is absolute, then in step ST61, the lower limit of the second data set needs to be adjusted (that is, the upper limit of the second data set does not need to be adjusted), so that the display of the two vertical axes is performed. The scale is the same, and the method is the second data set upper limit χ (the lower limit of the first data set + the first data set upper limit) in step ST611. In the second case, that is, after the absolute maximum 第一 of the first data group is greater than or equal to the minimum 値 absolute of the first data group, the maximum 値 of the second data group is absolute 値 less than the second data group. In the case where the minimum is absolutely absolute, then in step ST62, the upper limit of the second data set needs to be adjusted (that is, the lower limit of the second data set does not need to be adjusted) so that the display scales of the two vertical axes are the same, and the method thereof In step ST621, the second data set lower limit X (the lower limit of the first data set + the first data set upper limit). -14- (12) (12)200905490 In the third category, 'that is, after the absolute maximum of the first data set, the minimum is less than the minimum of the first data set, and the maximum of the second data set. After the absolute 値 is greater than or equal to the minimum 値 absolute 値 of the second data set, then in step ST63, the lower limit of the second data set needs to be adjusted (that is, the upper limit of the second data set does not need to be adjusted), so that the two vertical axes The display scale is the same 'the method is the second data set upper limit x (the upper limit of the first data set + the lower limit of the first data set) in step ST 6 31. In the fourth case, 'that is, after the absolute maximum of the first data set, the minimum is less than the minimum of the first data set. The maximum of the second data set is less than the second data set. In the case of the minimum absolute 値, the upper limit of the second data set needs to be adjusted (that is, the lower limit of the second data set does not need to be adjusted) in step ST64, so that the display scales of the two vertical axes are the same, by In step ST 6 4 1 , the second data set lower limit X (the upper limit of the first data set + the lower limit of the second data set). For example, there are now A and B sets of sequences to be compared with the two vertical axes: A series 5941 • 6382 7209 - 5323 7775 -3142 6117 5737 - 4657 7122 B series 72 84 92 96 71 82 85 69 76 82 When A, B series When the double vertical axis diagram is automatically created by EXCEL, the result as shown in Fig. 14a is obtained. However, from the 14b and 14c diagrams (both automatically drawn by EXCEL), it can be found that the double vertical axis diagram made by EXCEL only moves the left vertical axis of the original B-column single vertical axis automatic drawing to the right of the double vertical axis. The vertical axis, because the display scales of the two axes are not the same, put the two series on the same line graph 'can not be compared with each other fairly, and will exist: 1· Figure-15- (13) (13) 200905490 shape still shows the scale Non-maximum, the viewer is not easy to detect the trend, abnormality and major changes. 2. The proportional relationship between the two vertical axes is not considered, which makes it easy for the reader to misunderstand the fluctuation range of the two series, and this ratio problem is more likely Give the mapper the opportunity to manipulate the reader's feelings. Therefore, it is necessary to give a double column of the same scale to the vertical axis through the data processing flow of the proportional representation of the present invention. Since the E値 of the A series is 1.984823, and the E値 of the B series is 0.28125, the A series is the first data set, and the maximum and minimum A of the A series are still above and below the vertical axis, while the B series needs to be changed. The upper or lower limit of the longitudinal axis is such that the display scales of the two longitudinal axes of the two series are the same. Since the maximum of the A series is absolute, the minimum is greater than the minimum of the A series, and the maximum of the B series is greater than the minimum of the B series (belonging to the first case), therefore, B The lower limit of the vertical axis of the series needs to be corrected downward to achieve the same display scale of both sides, and the method is the upper limit x of the B number column (the lower limit of the A series and the upper limit of the A series). After adjusting the data processing method of the proportional representation method of the present invention, the upper limit of the A series is 7775, the lower limit of the A number is _63 82, the upper limit of the b number is 96, and the lower limit of the B number is -78.80025, as in the case of I4d. The figure shows. Through this method, you can get the maximum effect of two points: 1 · Because the vertical axis of the two data shows the same scale, the data can be displayed in a fairer proportion. 2. In the case where the two series shows equal scales, the normalized number is returned to its original unit number, making it easier for the reader to understand the size of the original data. The above fairness can be verified by comparing the numbers of the A and B series to 16 -16- (14) (14) 200905490 and comparing the 14e, 14f and 14d diagrams using the normalized representation. The number of normalized A and B series is as follows: A series 0.7641 -0.8208 0.9272 -0.6846 1 -0.4041 0.7867 0.7378 -0.5989 0.91601 B number column 0.75 0.875 0.9583 1 0.7395 0.8541 0.8854 0.7187 0.7916 0.8541 It can be seen from the above table that after normalization The maximum 値 of the A series is 1, the minimum 値 is -〇. 8 2 0 8 4, the maximum 値 of the B series is 1, and the minimum 0.7 is 0.7 1 87 5. The maximum and minimum values of the A and B series are normalized as the left and right axes of the double vertical axis, and the lower limit can be found that the vertical axis of the two displays the scale is completely different, so the comparison is not reasonable, such as the first It is shown in Fig. 4e; therefore, it is necessary to display the vertical axis to the smaller scale (that is, the maximum vertical axis and the minimum vertical axis of the upper and lower limits), and re-adjust to the larger scale with the vertical axis. Equivalent (that is, the maximum vertical axis of the A series and the left vertical axis of the lower limit), so that the maximum pull-up amplitude can be used to help the readers with the same ratio of the wheelbase ratios of the two. It is easy to detect trends and changes in numbers, as shown in Figure 14f. We can see that the map of Fig. 14f and the graph made by the proportional representation (Fig. 14d) are exactly the same, showing the fairness of the proportional representation, and the graph made by the equal proportion representation can make two The series returns to the original number, which helps the readers to more easily detect the actual size of the series. Of course, the present invention can also be represented by a line graph with a bar graph, and the data of the same graph of Fig. 14d can also be expressed as shown in Fig. 14g (the top and bottom sides are each doubled by the wheelbase width). -17- (15) (15)200905490 For example, there are now two sets of C and D series to do double vertical axis comparison · · C number column 72 84 92 96 71 82 85 69 76 82 D number column -1230 -4245 -2249 - 2654 -3234 -3790 -4593 -1983 -980 -789 When the C and D series automatically make a double vertical axis chart with EXCEL, you will get the result as shown in Figure 15a. However, it can be found from the 15b and 15c diagrams (both automatically drawn by EXCEL) that, similarly, the double longitudinal axis diagram made by EXCEL is to shift the left vertical axis of the original D-column automatic vertical drawing to the double vertical. The right vertical axis of the axis, because the display scales of the two axes are not the same, the two series are placed on the same line graph, which cannot be compared with each other fairly, and will exist: 1. The graph still shows that the scale is not the largest, the viewer does not It is easy to detect the trend, abnormality and major changes. 2. The proportional relationship between the two vertical axes is not considered, which makes it easy for the reader to misunderstand the fluctuation range of the two series. This ratio problem is more likely to be given to the mapper. The opportunity for people to feel. Therefore, it is necessary to give a double column of the same scale to the vertical axis through the data processing flow of the proportional representation of the present invention. Since the E 値 of the C number is 0.28125 and the E D of the D series is 0.828217', the 'D number is listed as the first data set. The maximum and minimum 値 of the D series are still above and below the vertical axis, while the C series needs to be changed. The upper or lower limit of the longitudinal axis is such that the display scales of the two longitudinal axes of the two series are the same. Since the maximum of the D number is absolutely less than the minimum of the D series, and the maximum of the C number is absolute, then the minimum is greater than the minimum of the c series (belonging to the third case), so 'C The vertical axis of the series -18- (16) (16) 200905490 needs to be corrected downward to achieve the same display scale of both sides. The method is the upper limit X of the C series (the upper limit of the D series - the lower limit of the D series). After adjusting by the proportional representation method of the present invention, the upper limit of the C number column is 96, the lower limit of the C number column is 16.491 18, the upper limit of the D number column is -7 8 9 , and the lower limit of the D number column is -4 5 9 3 , such as Figure 15D shows. Similarly, this method can get the maximum effect of two points: 1. Since the two data shows the same scale, the data can be displayed in a more fair scale. 2 • Under the condition that the vertical axis of the two series shows the same scale, the normalized number 値 is returned to its original unit number 値, making it easier for the reader to understand the size of the original data. The above fairness can also be verified by comparing the numbers of the C and D numbers and the 15e, 15f and 15d maps using the normalized representation. The number of normalized C-number and D-number series is as follows: C-number column 0.75 0.875 0.9583 1 0.7395 0.8541 0.8854 0.7187 0.7916 0.8541 D-number column -0.2678 -0.9242 -0.4896 -0.5778 -0.7041 -0.8251 -1 -0.4317 -0.2133 -0.1717 From the above table It can be seen that after normalization, the maximum 値 of the C-number column is 1, the minimum 値 is 0.7 1 8 7 5, the maximum D of the D-number column is _0.1 7 1 784, and the minimum 値 is -1. The maximum and minimum values of the C and D series are normalized as the left and right axes of the double vertical axis diagram. The lower limit can be found that the wheelbase display scales of the two are completely different, so the comparison is not reasonable, such as the first 5 e picture; therefore, it is necessary to re-adjust the side with the smaller scale of the wheelbase display (that is, the left vertical axis with the maximum and minimum C of the C number) and the vertical axis -19- ( 17) (17) 200905490 The one with the larger scale is equal (that is, the right vertical axis with the maximum and minimum D of the D number). Thus, the ratio of the vertical axis of the two can be achieved. With the maximum pull-up amplitude, the reader is most likely to perceive the trend and change of the number, as shown in Figure 15f. We can see that the figure 15f and the figure made by the proportional representation (Fig. 15d) are exactly the same as the 'fairness of the equal-representation method'. The return of the two series to the original number 値 'help readers are more likely to detect the size of the original data. Of course, the present invention can also be represented by a line graph with a bar graph. The data similar to the 15d graph can also be expressed as shown in Fig. 15g (the upper and lower sides are doubled by the wheelbase width). To sum up, this case is a method for graphing data, which can aggregate and analyze a large amount of data' and use the actual 値 representation disclosed in this case according to the different needs of the readers and the number of data. The normalized representation, or the different data processing methods of the proportional representation method, so that the reader can more clearly know the substantive meaning of the data, and will not be manipulated by the mapper. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart for explaining the steps of graphing data; FIG. 2 is a schematic diagram for illustrating further details of the steps of processing data in FIG. 1; For a flow chart, to explain the actual 値 representation processing -20- (18) (18) 200905490 material steps; Figure 4 is a flow chart to explain the formalization table two methods of processing data; 5 is a flow chart for explaining the processing data step of the proportional representation method; FIG. 6 is a flow chart for explaining further details of the step of correcting the upper and lower limits of the equalization table; 7a. 7b, 8a, 8b are used to illustrate examples of different drawing methods that give the viewer a completely different experience; Figures 9a and 9b are single vertical axis singular numbers obtained using the traditional method and the actual 値 representation of the present invention, respectively. Column line chart; Figures 10a and 10b are respectively a single vertical axis singular column line diagram obtained by using the conventional method and the actual 値 representation of the present invention plus the average 値 line and 〇 reference line: The 11th and lib diagrams respectively use the conventional Method, the reality of the present invention单 値 之 单 ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; The method, the actual 値 representation of the present invention, the single vertical axis majority column line graph obtained by the normalized representation of the present invention; the 13a, 13b, and 13c graphs respectively use the conventional mapping method and the present invention to find the series The maximum and minimum 値 are used as the normalized representation of the upper and lower limits, the upper limit of the setting of the present invention is 1, and the lower limit is the regularized representation - 21 (19) (19) 200905490. The figures i 4a, 14b, and 14c are the double longitudinal axis diagrams of the A and B series obtained by the conventional mapping method, the single vertical axis diagram of the A series, and the single vertical axis diagram of the B series; the first 4th figure is The double longitudinal axis map of the A and B series obtained by the proportional representation method of the present invention; the first and fourth 4 f f maps respectively use the maximum and minimum 値 after normalization as the upper and lower limits of the two longitudinal axes respectively. The figure and the regularization of the normalization of the double vertical axis The linear graph; the first 4 g graph is a double longitudinal axis-shaped bar graph of the A and B series obtained by using the proportional representation method of the present invention; the 15th, 15th, and 15th graphs are respectively used The double vertical axis graph of the C and D numbers, the single vertical axis graph of the C number sequence, and the single vertical axis graph of the D number sequence obtained by the conventional drawing method; the 15th graph is obtained by using the proportional representation method of the present invention. The double longitudinal axis map of the D series; the 15th and 15th f maps respectively use the maximum and minimum 正规 after normalization as the 'lower limit' on the two vertical axes and the double vertical axis to be normalized. The line graph obtained after the proportional correction; and the 15th graph are the double longitudinal axis-shaped bar graphs of the C and D numbers obtained by using the proportional representation method of the present invention. -twenty two-